Embed Size (px)
Citation preview
Adoption of Re ommended Maize Technologies middotn Mbeya egion of
the Southern Highlands of Tanzania
Shekania Bisanda and
Wilfred Mwangt
The United Republic of Tanzallla MInistry of Agriculture
CIMMYTfEU Proj~ct on Strengthening Economic and Polky Research in ARS In Eastern Africa
Adoption of Recommended Maize Technologies in Mbeya Region of
the Southern Highlands of Tanzania
Shekania Bisanda FSR Sociologist MARTI Uyole PO Box 400 Mbeya Tanzania
and
Wilfred Mwangi Regional Economist CIMMYT Po Box 5689 Addis Ababa Ethiopia
The United Republic of Thnzania Ministry of Agriculture
International Maize and Wheat Improvement Center (CIMMYT )
Published by the International Maire and Wheat Improvement Center PO Box 5689 Addis Ababa and The United Republic of Tanzania Ministry of Agriculture Po Box 2066 Dar es salam Tanzania
Published 1996
Printed at ILRI Addis Ababa Ethiopia
Correct citation Shekania Bisanda and Wilfred Mwangi 1996 Adoption of recomshymended maize in of the southern highlands of Tanzania Addis Ababa CIMMYTthe United of Tanzania Ministry of Agriculture
iJ in this paper are those of the authors and should not be attributed to
ISBN 92~9146~013-3
Contents
List of tables IV
List of figures V
Acknowledgements Vi
Summary vii
Introduction
Methodology 3
Description of the study area 5
Maize research in the southern highlands 8
Socioeconomic and demographic characteristics of furmers 14
Adoption of recommendations 22
Factors affecting adoption 36
Conclusions and implications fur research extension and policy 38
References 40
iii
List of tables
Table I Household head by gender and district Table 2 Characteristics of released improved maize varieties Table 3 Recommended maize varieties for different wards and agro-ecological
zones Table 4 Recommended spacing and seeding rate Table 5 Effect of different weeding regimes on maize grain yield Table 6 A summary of recommended practices for maize production Table 7 Potential yield (tha) of high and mid-altitude maize cuitivars III
different locations in the Southern Highlands Table 8 Range of years lived in the village by gender of household heads Table 9 Farm size and cultivated area by gender of household head Table 10 Area under various crops (in acres) by gender of household head Table II Number of maize plots by gender of household head Table 12 Reasons for fullowing land by gender of household head Table 13 Hired labor used for maize furming by gender of household head Table 14 Mean number of cattle and other livestock kept by gender of
household head Table 15 Time of land preparation by plot of land and gender of household
head Table 16 Spacing between rows by gender of household head Table 17 Row planting by year of adoption and gender of household head Table 18 Intercropping maize with other crops by gender of household head Table 19 Growing of improved maize variety by gender of household head Table 20 Improved maize varieties ever grown by gender of household head Table 21 Period when furmers started growing improved maize variety by
gender of household head Table 22 Period improved maize variety was discarded by gender of household
head Table 23 Date of first weeding of plot one by gender of household head Table 24 Reasons for not using herbicide by gender of household head Table 25 Reasons for not using fertilizers by gender of household head Table 26 Period when furmers started to use inorganic fertilizer by gender of
household head Table 27 Distances travelled to obtain organic and inorganic fertilizers by
gender of household head Table 28 Maize storage systems by gender of household head Table 29 Sources of infurmation for obtaining hybrid maize seeds by gender of
household head
iv
List of figures
Figure 1 Thnzania The Location of the four regions of the Southern Highlands Zone
Figure 2 Mbeya Region The Location of the Study Area
v
ACKNOWLEDGEMENTS
The authors of this report thank the Commissioner of Research and Training Ministry of Agriculture Thnzania and CIMMYT fur providing financial and institutional support
We thank Mr TN Kirway the Assistant Commissioner fur Farming Systems Research in the Ministry of Agriculture (Tanzania) fur facilitating this study and taking part in pre-testing the questionnaire Also we are thankful to Christian Mwakyembe Zakaria Malley Mathias Mkoma Francis Mosses Arnan Hallinga and Alatwisila Sakalani fur administering the questionnaire We are grateful to breeders and agronomists in the Maize Improvement Program at Uyole fur their technical input
We acknowledge the contribution of furmers in the study area fur responding to our questions willingly The contribution of the district and village level extension workers in all the three districts with regard to this report is highly appreciated
Last but not least we are greatly indebted to the secretaries Yohana Lukindo Issabella Minga and Aklilewerk Bekele fur expediting the publication of this report
VI
SUMMARY
Background to the study
This study was undertaken in order to document the level and rate of adoption of improved maize technologies in Mbeya region in the Southern Highlands of Tanzania The study had the following objectives
bull to investigate the rate and sequence of adoption bull to examine some characteristics of adopting and non-adopting fumers bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
Methodology
Initially literature review was made in order to get acquainted with research and extension services in the study area Based on literature review a reconnaissance survey was conducted in order to get first-hand infunnation about the study area Thereafter a questionnaire was administered to 100 funners in Mbeya Rungwe and Mbozi districts Fanners were either purposively or randomly selected depending on the required infunnation Multiple visits were made to each sample funner after each funn operation to capture their actual practices
Description of the study area
The study villages were from four different agro-ecological zones which practice dry season planting These included the Central Mbeya zone (1200 m) Mbozi Isuto and Ndola plateau and Poroto mountains (2400 m) and Tukuyu volcano and Rungwe valleys (2400 m)
Maize research in the Southern Highlands
Maize research in Southern Highlands started during the 1970171 cropping season at Uyole Agricultural Center (now MARTI Uyole) in Mbeya Municipality under the Tanzania Nordic Agricultural Project These activities increased steadily after 1974 following the establishment of the National Maize Research Program (NMRP) In 1981 maize research at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) The MIP was charged to develop appropriate varieties to improve parental lines of currently used commercial hybrids and to research on agronomic practices pest and disease control Recommended varieties in the study area include H614 UCA Kilima TMV-l and TMV-2
vii
Socio-economic characteristics of farmers
The majority (90) of male-headed households had almost 5 persons and about 42 of them had only 2 persons The average number of adult household members for female-headed fumilies was 24 persons Household size varied more in femaleshyheaded households than in the male ones The mean household size for male-headed families was significantly different from that of female-headed households More adults from male-headed fumilies worked on farms pennanently than the number of adults from female-headed households Most male family heads (96 ) had completed elementary fonnal education (below 7 years) while most female family heads (82 ) never went to school
Land tenure
Land ownership is communal and fanners have individual user rights Most male heads (825 ) and female heads (973 ) occupied a fann size of less than 10 acres Maize beans coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively Maize was the most important crop as it was cultivated by all respondents Seventy percent of male- and female-headed households did not tallow land Those who fallowed did so to increase soil fertility
Farm mechanization
Sixty-seven percent of all respondents used hand hoes and only 11 used oxen for cultivation and 22 used both hand hoes and oxen Most female fumily heads tended to use hand hoes (none used oxen) About balf of the surveyed households used hired labor mainly for land preparation wbile 21 hired labor for weeding and harvesting In both cases more female-beaded households hired labor than the male-headed households
Adoption of recommendations
Most fanners (70 ) started land preparation between July and November depending on the planting season i e dry or wet season Hovever where dry season planting was practiced plowing and planting were done concurrently About 57 of maleshyheaded households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
Planting practices
Almost all fanners adopted row planting Most female-headed households adopted row planting between 1986 and 1990 while male-headed households adopted before this period None of the sampled fanners follow the recommended levels of spacing but the majority who used 30 x 90 cm were very close to the recommended spacing
viii
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Adoption of Recommended Maize Technologies in Mbeya Region of
the Southern Highlands of Tanzania
Shekania Bisanda FSR Sociologist MARTI Uyole PO Box 400 Mbeya Tanzania
and
Wilfred Mwangi Regional Economist CIMMYT Po Box 5689 Addis Ababa Ethiopia
The United Republic of Thnzania Ministry of Agriculture
International Maize and Wheat Improvement Center (CIMMYT )
Published by the International Maire and Wheat Improvement Center PO Box 5689 Addis Ababa and The United Republic of Tanzania Ministry of Agriculture Po Box 2066 Dar es salam Tanzania
Published 1996
Printed at ILRI Addis Ababa Ethiopia
Correct citation Shekania Bisanda and Wilfred Mwangi 1996 Adoption of recomshymended maize in of the southern highlands of Tanzania Addis Ababa CIMMYTthe United of Tanzania Ministry of Agriculture
iJ in this paper are those of the authors and should not be attributed to
ISBN 92~9146~013-3
Contents
List of tables IV
List of figures V
Acknowledgements Vi
Summary vii
Introduction
Methodology 3
Description of the study area 5
Maize research in the southern highlands 8
Socioeconomic and demographic characteristics of furmers 14
Adoption of recommendations 22
Factors affecting adoption 36
Conclusions and implications fur research extension and policy 38
References 40
iii
List of tables
Table I Household head by gender and district Table 2 Characteristics of released improved maize varieties Table 3 Recommended maize varieties for different wards and agro-ecological
zones Table 4 Recommended spacing and seeding rate Table 5 Effect of different weeding regimes on maize grain yield Table 6 A summary of recommended practices for maize production Table 7 Potential yield (tha) of high and mid-altitude maize cuitivars III
different locations in the Southern Highlands Table 8 Range of years lived in the village by gender of household heads Table 9 Farm size and cultivated area by gender of household head Table 10 Area under various crops (in acres) by gender of household head Table II Number of maize plots by gender of household head Table 12 Reasons for fullowing land by gender of household head Table 13 Hired labor used for maize furming by gender of household head Table 14 Mean number of cattle and other livestock kept by gender of
household head Table 15 Time of land preparation by plot of land and gender of household
head Table 16 Spacing between rows by gender of household head Table 17 Row planting by year of adoption and gender of household head Table 18 Intercropping maize with other crops by gender of household head Table 19 Growing of improved maize variety by gender of household head Table 20 Improved maize varieties ever grown by gender of household head Table 21 Period when furmers started growing improved maize variety by
gender of household head Table 22 Period improved maize variety was discarded by gender of household
head Table 23 Date of first weeding of plot one by gender of household head Table 24 Reasons for not using herbicide by gender of household head Table 25 Reasons for not using fertilizers by gender of household head Table 26 Period when furmers started to use inorganic fertilizer by gender of
household head Table 27 Distances travelled to obtain organic and inorganic fertilizers by
gender of household head Table 28 Maize storage systems by gender of household head Table 29 Sources of infurmation for obtaining hybrid maize seeds by gender of
household head
iv
List of figures
Figure 1 Thnzania The Location of the four regions of the Southern Highlands Zone
Figure 2 Mbeya Region The Location of the Study Area
v
ACKNOWLEDGEMENTS
The authors of this report thank the Commissioner of Research and Training Ministry of Agriculture Thnzania and CIMMYT fur providing financial and institutional support
We thank Mr TN Kirway the Assistant Commissioner fur Farming Systems Research in the Ministry of Agriculture (Tanzania) fur facilitating this study and taking part in pre-testing the questionnaire Also we are thankful to Christian Mwakyembe Zakaria Malley Mathias Mkoma Francis Mosses Arnan Hallinga and Alatwisila Sakalani fur administering the questionnaire We are grateful to breeders and agronomists in the Maize Improvement Program at Uyole fur their technical input
We acknowledge the contribution of furmers in the study area fur responding to our questions willingly The contribution of the district and village level extension workers in all the three districts with regard to this report is highly appreciated
Last but not least we are greatly indebted to the secretaries Yohana Lukindo Issabella Minga and Aklilewerk Bekele fur expediting the publication of this report
VI
SUMMARY
Background to the study
This study was undertaken in order to document the level and rate of adoption of improved maize technologies in Mbeya region in the Southern Highlands of Tanzania The study had the following objectives
bull to investigate the rate and sequence of adoption bull to examine some characteristics of adopting and non-adopting fumers bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
Methodology
Initially literature review was made in order to get acquainted with research and extension services in the study area Based on literature review a reconnaissance survey was conducted in order to get first-hand infunnation about the study area Thereafter a questionnaire was administered to 100 funners in Mbeya Rungwe and Mbozi districts Fanners were either purposively or randomly selected depending on the required infunnation Multiple visits were made to each sample funner after each funn operation to capture their actual practices
Description of the study area
The study villages were from four different agro-ecological zones which practice dry season planting These included the Central Mbeya zone (1200 m) Mbozi Isuto and Ndola plateau and Poroto mountains (2400 m) and Tukuyu volcano and Rungwe valleys (2400 m)
Maize research in the Southern Highlands
Maize research in Southern Highlands started during the 1970171 cropping season at Uyole Agricultural Center (now MARTI Uyole) in Mbeya Municipality under the Tanzania Nordic Agricultural Project These activities increased steadily after 1974 following the establishment of the National Maize Research Program (NMRP) In 1981 maize research at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) The MIP was charged to develop appropriate varieties to improve parental lines of currently used commercial hybrids and to research on agronomic practices pest and disease control Recommended varieties in the study area include H614 UCA Kilima TMV-l and TMV-2
vii
Socio-economic characteristics of farmers
The majority (90) of male-headed households had almost 5 persons and about 42 of them had only 2 persons The average number of adult household members for female-headed fumilies was 24 persons Household size varied more in femaleshyheaded households than in the male ones The mean household size for male-headed families was significantly different from that of female-headed households More adults from male-headed fumilies worked on farms pennanently than the number of adults from female-headed households Most male family heads (96 ) had completed elementary fonnal education (below 7 years) while most female family heads (82 ) never went to school
Land tenure
Land ownership is communal and fanners have individual user rights Most male heads (825 ) and female heads (973 ) occupied a fann size of less than 10 acres Maize beans coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively Maize was the most important crop as it was cultivated by all respondents Seventy percent of male- and female-headed households did not tallow land Those who fallowed did so to increase soil fertility
Farm mechanization
Sixty-seven percent of all respondents used hand hoes and only 11 used oxen for cultivation and 22 used both hand hoes and oxen Most female fumily heads tended to use hand hoes (none used oxen) About balf of the surveyed households used hired labor mainly for land preparation wbile 21 hired labor for weeding and harvesting In both cases more female-beaded households hired labor than the male-headed households
Adoption of recommendations
Most fanners (70 ) started land preparation between July and November depending on the planting season i e dry or wet season Hovever where dry season planting was practiced plowing and planting were done concurrently About 57 of maleshyheaded households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
Planting practices
Almost all fanners adopted row planting Most female-headed households adopted row planting between 1986 and 1990 while male-headed households adopted before this period None of the sampled fanners follow the recommended levels of spacing but the majority who used 30 x 90 cm were very close to the recommended spacing
viii
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Published by the International Maire and Wheat Improvement Center PO Box 5689 Addis Ababa and The United Republic of Tanzania Ministry of Agriculture Po Box 2066 Dar es salam Tanzania
Published 1996
Printed at ILRI Addis Ababa Ethiopia
Correct citation Shekania Bisanda and Wilfred Mwangi 1996 Adoption of recomshymended maize in of the southern highlands of Tanzania Addis Ababa CIMMYTthe United of Tanzania Ministry of Agriculture
iJ in this paper are those of the authors and should not be attributed to
ISBN 92~9146~013-3
Contents
List of tables IV
List of figures V
Acknowledgements Vi
Summary vii
Introduction
Methodology 3
Description of the study area 5
Maize research in the southern highlands 8
Socioeconomic and demographic characteristics of furmers 14
Adoption of recommendations 22
Factors affecting adoption 36
Conclusions and implications fur research extension and policy 38
References 40
iii
List of tables
Table I Household head by gender and district Table 2 Characteristics of released improved maize varieties Table 3 Recommended maize varieties for different wards and agro-ecological
zones Table 4 Recommended spacing and seeding rate Table 5 Effect of different weeding regimes on maize grain yield Table 6 A summary of recommended practices for maize production Table 7 Potential yield (tha) of high and mid-altitude maize cuitivars III
different locations in the Southern Highlands Table 8 Range of years lived in the village by gender of household heads Table 9 Farm size and cultivated area by gender of household head Table 10 Area under various crops (in acres) by gender of household head Table II Number of maize plots by gender of household head Table 12 Reasons for fullowing land by gender of household head Table 13 Hired labor used for maize furming by gender of household head Table 14 Mean number of cattle and other livestock kept by gender of
household head Table 15 Time of land preparation by plot of land and gender of household
head Table 16 Spacing between rows by gender of household head Table 17 Row planting by year of adoption and gender of household head Table 18 Intercropping maize with other crops by gender of household head Table 19 Growing of improved maize variety by gender of household head Table 20 Improved maize varieties ever grown by gender of household head Table 21 Period when furmers started growing improved maize variety by
gender of household head Table 22 Period improved maize variety was discarded by gender of household
head Table 23 Date of first weeding of plot one by gender of household head Table 24 Reasons for not using herbicide by gender of household head Table 25 Reasons for not using fertilizers by gender of household head Table 26 Period when furmers started to use inorganic fertilizer by gender of
household head Table 27 Distances travelled to obtain organic and inorganic fertilizers by
gender of household head Table 28 Maize storage systems by gender of household head Table 29 Sources of infurmation for obtaining hybrid maize seeds by gender of
household head
iv
List of figures
Figure 1 Thnzania The Location of the four regions of the Southern Highlands Zone
Figure 2 Mbeya Region The Location of the Study Area
v
ACKNOWLEDGEMENTS
The authors of this report thank the Commissioner of Research and Training Ministry of Agriculture Thnzania and CIMMYT fur providing financial and institutional support
We thank Mr TN Kirway the Assistant Commissioner fur Farming Systems Research in the Ministry of Agriculture (Tanzania) fur facilitating this study and taking part in pre-testing the questionnaire Also we are thankful to Christian Mwakyembe Zakaria Malley Mathias Mkoma Francis Mosses Arnan Hallinga and Alatwisila Sakalani fur administering the questionnaire We are grateful to breeders and agronomists in the Maize Improvement Program at Uyole fur their technical input
We acknowledge the contribution of furmers in the study area fur responding to our questions willingly The contribution of the district and village level extension workers in all the three districts with regard to this report is highly appreciated
Last but not least we are greatly indebted to the secretaries Yohana Lukindo Issabella Minga and Aklilewerk Bekele fur expediting the publication of this report
VI
SUMMARY
Background to the study
This study was undertaken in order to document the level and rate of adoption of improved maize technologies in Mbeya region in the Southern Highlands of Tanzania The study had the following objectives
bull to investigate the rate and sequence of adoption bull to examine some characteristics of adopting and non-adopting fumers bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
Methodology
Initially literature review was made in order to get acquainted with research and extension services in the study area Based on literature review a reconnaissance survey was conducted in order to get first-hand infunnation about the study area Thereafter a questionnaire was administered to 100 funners in Mbeya Rungwe and Mbozi districts Fanners were either purposively or randomly selected depending on the required infunnation Multiple visits were made to each sample funner after each funn operation to capture their actual practices
Description of the study area
The study villages were from four different agro-ecological zones which practice dry season planting These included the Central Mbeya zone (1200 m) Mbozi Isuto and Ndola plateau and Poroto mountains (2400 m) and Tukuyu volcano and Rungwe valleys (2400 m)
Maize research in the Southern Highlands
Maize research in Southern Highlands started during the 1970171 cropping season at Uyole Agricultural Center (now MARTI Uyole) in Mbeya Municipality under the Tanzania Nordic Agricultural Project These activities increased steadily after 1974 following the establishment of the National Maize Research Program (NMRP) In 1981 maize research at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) The MIP was charged to develop appropriate varieties to improve parental lines of currently used commercial hybrids and to research on agronomic practices pest and disease control Recommended varieties in the study area include H614 UCA Kilima TMV-l and TMV-2
vii
Socio-economic characteristics of farmers
The majority (90) of male-headed households had almost 5 persons and about 42 of them had only 2 persons The average number of adult household members for female-headed fumilies was 24 persons Household size varied more in femaleshyheaded households than in the male ones The mean household size for male-headed families was significantly different from that of female-headed households More adults from male-headed fumilies worked on farms pennanently than the number of adults from female-headed households Most male family heads (96 ) had completed elementary fonnal education (below 7 years) while most female family heads (82 ) never went to school
Land tenure
Land ownership is communal and fanners have individual user rights Most male heads (825 ) and female heads (973 ) occupied a fann size of less than 10 acres Maize beans coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively Maize was the most important crop as it was cultivated by all respondents Seventy percent of male- and female-headed households did not tallow land Those who fallowed did so to increase soil fertility
Farm mechanization
Sixty-seven percent of all respondents used hand hoes and only 11 used oxen for cultivation and 22 used both hand hoes and oxen Most female fumily heads tended to use hand hoes (none used oxen) About balf of the surveyed households used hired labor mainly for land preparation wbile 21 hired labor for weeding and harvesting In both cases more female-beaded households hired labor than the male-headed households
Adoption of recommendations
Most fanners (70 ) started land preparation between July and November depending on the planting season i e dry or wet season Hovever where dry season planting was practiced plowing and planting were done concurrently About 57 of maleshyheaded households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
Planting practices
Almost all fanners adopted row planting Most female-headed households adopted row planting between 1986 and 1990 while male-headed households adopted before this period None of the sampled fanners follow the recommended levels of spacing but the majority who used 30 x 90 cm were very close to the recommended spacing
viii
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Contents
List of tables IV
List of figures V
Acknowledgements Vi
Summary vii
Introduction
Methodology 3
Description of the study area 5
Maize research in the southern highlands 8
Socioeconomic and demographic characteristics of furmers 14
Adoption of recommendations 22
Factors affecting adoption 36
Conclusions and implications fur research extension and policy 38
References 40
iii
List of tables
Table I Household head by gender and district Table 2 Characteristics of released improved maize varieties Table 3 Recommended maize varieties for different wards and agro-ecological
zones Table 4 Recommended spacing and seeding rate Table 5 Effect of different weeding regimes on maize grain yield Table 6 A summary of recommended practices for maize production Table 7 Potential yield (tha) of high and mid-altitude maize cuitivars III
different locations in the Southern Highlands Table 8 Range of years lived in the village by gender of household heads Table 9 Farm size and cultivated area by gender of household head Table 10 Area under various crops (in acres) by gender of household head Table II Number of maize plots by gender of household head Table 12 Reasons for fullowing land by gender of household head Table 13 Hired labor used for maize furming by gender of household head Table 14 Mean number of cattle and other livestock kept by gender of
household head Table 15 Time of land preparation by plot of land and gender of household
head Table 16 Spacing between rows by gender of household head Table 17 Row planting by year of adoption and gender of household head Table 18 Intercropping maize with other crops by gender of household head Table 19 Growing of improved maize variety by gender of household head Table 20 Improved maize varieties ever grown by gender of household head Table 21 Period when furmers started growing improved maize variety by
gender of household head Table 22 Period improved maize variety was discarded by gender of household
head Table 23 Date of first weeding of plot one by gender of household head Table 24 Reasons for not using herbicide by gender of household head Table 25 Reasons for not using fertilizers by gender of household head Table 26 Period when furmers started to use inorganic fertilizer by gender of
household head Table 27 Distances travelled to obtain organic and inorganic fertilizers by
gender of household head Table 28 Maize storage systems by gender of household head Table 29 Sources of infurmation for obtaining hybrid maize seeds by gender of
household head
iv
List of figures
Figure 1 Thnzania The Location of the four regions of the Southern Highlands Zone
Figure 2 Mbeya Region The Location of the Study Area
v
ACKNOWLEDGEMENTS
The authors of this report thank the Commissioner of Research and Training Ministry of Agriculture Thnzania and CIMMYT fur providing financial and institutional support
We thank Mr TN Kirway the Assistant Commissioner fur Farming Systems Research in the Ministry of Agriculture (Tanzania) fur facilitating this study and taking part in pre-testing the questionnaire Also we are thankful to Christian Mwakyembe Zakaria Malley Mathias Mkoma Francis Mosses Arnan Hallinga and Alatwisila Sakalani fur administering the questionnaire We are grateful to breeders and agronomists in the Maize Improvement Program at Uyole fur their technical input
We acknowledge the contribution of furmers in the study area fur responding to our questions willingly The contribution of the district and village level extension workers in all the three districts with regard to this report is highly appreciated
Last but not least we are greatly indebted to the secretaries Yohana Lukindo Issabella Minga and Aklilewerk Bekele fur expediting the publication of this report
VI
SUMMARY
Background to the study
This study was undertaken in order to document the level and rate of adoption of improved maize technologies in Mbeya region in the Southern Highlands of Tanzania The study had the following objectives
bull to investigate the rate and sequence of adoption bull to examine some characteristics of adopting and non-adopting fumers bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
Methodology
Initially literature review was made in order to get acquainted with research and extension services in the study area Based on literature review a reconnaissance survey was conducted in order to get first-hand infunnation about the study area Thereafter a questionnaire was administered to 100 funners in Mbeya Rungwe and Mbozi districts Fanners were either purposively or randomly selected depending on the required infunnation Multiple visits were made to each sample funner after each funn operation to capture their actual practices
Description of the study area
The study villages were from four different agro-ecological zones which practice dry season planting These included the Central Mbeya zone (1200 m) Mbozi Isuto and Ndola plateau and Poroto mountains (2400 m) and Tukuyu volcano and Rungwe valleys (2400 m)
Maize research in the Southern Highlands
Maize research in Southern Highlands started during the 1970171 cropping season at Uyole Agricultural Center (now MARTI Uyole) in Mbeya Municipality under the Tanzania Nordic Agricultural Project These activities increased steadily after 1974 following the establishment of the National Maize Research Program (NMRP) In 1981 maize research at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) The MIP was charged to develop appropriate varieties to improve parental lines of currently used commercial hybrids and to research on agronomic practices pest and disease control Recommended varieties in the study area include H614 UCA Kilima TMV-l and TMV-2
vii
Socio-economic characteristics of farmers
The majority (90) of male-headed households had almost 5 persons and about 42 of them had only 2 persons The average number of adult household members for female-headed fumilies was 24 persons Household size varied more in femaleshyheaded households than in the male ones The mean household size for male-headed families was significantly different from that of female-headed households More adults from male-headed fumilies worked on farms pennanently than the number of adults from female-headed households Most male family heads (96 ) had completed elementary fonnal education (below 7 years) while most female family heads (82 ) never went to school
Land tenure
Land ownership is communal and fanners have individual user rights Most male heads (825 ) and female heads (973 ) occupied a fann size of less than 10 acres Maize beans coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively Maize was the most important crop as it was cultivated by all respondents Seventy percent of male- and female-headed households did not tallow land Those who fallowed did so to increase soil fertility
Farm mechanization
Sixty-seven percent of all respondents used hand hoes and only 11 used oxen for cultivation and 22 used both hand hoes and oxen Most female fumily heads tended to use hand hoes (none used oxen) About balf of the surveyed households used hired labor mainly for land preparation wbile 21 hired labor for weeding and harvesting In both cases more female-beaded households hired labor than the male-headed households
Adoption of recommendations
Most fanners (70 ) started land preparation between July and November depending on the planting season i e dry or wet season Hovever where dry season planting was practiced plowing and planting were done concurrently About 57 of maleshyheaded households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
Planting practices
Almost all fanners adopted row planting Most female-headed households adopted row planting between 1986 and 1990 while male-headed households adopted before this period None of the sampled fanners follow the recommended levels of spacing but the majority who used 30 x 90 cm were very close to the recommended spacing
viii
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
List of tables
Table I Household head by gender and district Table 2 Characteristics of released improved maize varieties Table 3 Recommended maize varieties for different wards and agro-ecological
zones Table 4 Recommended spacing and seeding rate Table 5 Effect of different weeding regimes on maize grain yield Table 6 A summary of recommended practices for maize production Table 7 Potential yield (tha) of high and mid-altitude maize cuitivars III
different locations in the Southern Highlands Table 8 Range of years lived in the village by gender of household heads Table 9 Farm size and cultivated area by gender of household head Table 10 Area under various crops (in acres) by gender of household head Table II Number of maize plots by gender of household head Table 12 Reasons for fullowing land by gender of household head Table 13 Hired labor used for maize furming by gender of household head Table 14 Mean number of cattle and other livestock kept by gender of
household head Table 15 Time of land preparation by plot of land and gender of household
head Table 16 Spacing between rows by gender of household head Table 17 Row planting by year of adoption and gender of household head Table 18 Intercropping maize with other crops by gender of household head Table 19 Growing of improved maize variety by gender of household head Table 20 Improved maize varieties ever grown by gender of household head Table 21 Period when furmers started growing improved maize variety by
gender of household head Table 22 Period improved maize variety was discarded by gender of household
head Table 23 Date of first weeding of plot one by gender of household head Table 24 Reasons for not using herbicide by gender of household head Table 25 Reasons for not using fertilizers by gender of household head Table 26 Period when furmers started to use inorganic fertilizer by gender of
household head Table 27 Distances travelled to obtain organic and inorganic fertilizers by
gender of household head Table 28 Maize storage systems by gender of household head Table 29 Sources of infurmation for obtaining hybrid maize seeds by gender of
household head
iv
List of figures
Figure 1 Thnzania The Location of the four regions of the Southern Highlands Zone
Figure 2 Mbeya Region The Location of the Study Area
v
ACKNOWLEDGEMENTS
The authors of this report thank the Commissioner of Research and Training Ministry of Agriculture Thnzania and CIMMYT fur providing financial and institutional support
We thank Mr TN Kirway the Assistant Commissioner fur Farming Systems Research in the Ministry of Agriculture (Tanzania) fur facilitating this study and taking part in pre-testing the questionnaire Also we are thankful to Christian Mwakyembe Zakaria Malley Mathias Mkoma Francis Mosses Arnan Hallinga and Alatwisila Sakalani fur administering the questionnaire We are grateful to breeders and agronomists in the Maize Improvement Program at Uyole fur their technical input
We acknowledge the contribution of furmers in the study area fur responding to our questions willingly The contribution of the district and village level extension workers in all the three districts with regard to this report is highly appreciated
Last but not least we are greatly indebted to the secretaries Yohana Lukindo Issabella Minga and Aklilewerk Bekele fur expediting the publication of this report
VI
SUMMARY
Background to the study
This study was undertaken in order to document the level and rate of adoption of improved maize technologies in Mbeya region in the Southern Highlands of Tanzania The study had the following objectives
bull to investigate the rate and sequence of adoption bull to examine some characteristics of adopting and non-adopting fumers bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
Methodology
Initially literature review was made in order to get acquainted with research and extension services in the study area Based on literature review a reconnaissance survey was conducted in order to get first-hand infunnation about the study area Thereafter a questionnaire was administered to 100 funners in Mbeya Rungwe and Mbozi districts Fanners were either purposively or randomly selected depending on the required infunnation Multiple visits were made to each sample funner after each funn operation to capture their actual practices
Description of the study area
The study villages were from four different agro-ecological zones which practice dry season planting These included the Central Mbeya zone (1200 m) Mbozi Isuto and Ndola plateau and Poroto mountains (2400 m) and Tukuyu volcano and Rungwe valleys (2400 m)
Maize research in the Southern Highlands
Maize research in Southern Highlands started during the 1970171 cropping season at Uyole Agricultural Center (now MARTI Uyole) in Mbeya Municipality under the Tanzania Nordic Agricultural Project These activities increased steadily after 1974 following the establishment of the National Maize Research Program (NMRP) In 1981 maize research at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) The MIP was charged to develop appropriate varieties to improve parental lines of currently used commercial hybrids and to research on agronomic practices pest and disease control Recommended varieties in the study area include H614 UCA Kilima TMV-l and TMV-2
vii
Socio-economic characteristics of farmers
The majority (90) of male-headed households had almost 5 persons and about 42 of them had only 2 persons The average number of adult household members for female-headed fumilies was 24 persons Household size varied more in femaleshyheaded households than in the male ones The mean household size for male-headed families was significantly different from that of female-headed households More adults from male-headed fumilies worked on farms pennanently than the number of adults from female-headed households Most male family heads (96 ) had completed elementary fonnal education (below 7 years) while most female family heads (82 ) never went to school
Land tenure
Land ownership is communal and fanners have individual user rights Most male heads (825 ) and female heads (973 ) occupied a fann size of less than 10 acres Maize beans coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively Maize was the most important crop as it was cultivated by all respondents Seventy percent of male- and female-headed households did not tallow land Those who fallowed did so to increase soil fertility
Farm mechanization
Sixty-seven percent of all respondents used hand hoes and only 11 used oxen for cultivation and 22 used both hand hoes and oxen Most female fumily heads tended to use hand hoes (none used oxen) About balf of the surveyed households used hired labor mainly for land preparation wbile 21 hired labor for weeding and harvesting In both cases more female-beaded households hired labor than the male-headed households
Adoption of recommendations
Most fanners (70 ) started land preparation between July and November depending on the planting season i e dry or wet season Hovever where dry season planting was practiced plowing and planting were done concurrently About 57 of maleshyheaded households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
Planting practices
Almost all fanners adopted row planting Most female-headed households adopted row planting between 1986 and 1990 while male-headed households adopted before this period None of the sampled fanners follow the recommended levels of spacing but the majority who used 30 x 90 cm were very close to the recommended spacing
viii
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
List of figures
Figure 1 Thnzania The Location of the four regions of the Southern Highlands Zone
Figure 2 Mbeya Region The Location of the Study Area
v
ACKNOWLEDGEMENTS
The authors of this report thank the Commissioner of Research and Training Ministry of Agriculture Thnzania and CIMMYT fur providing financial and institutional support
We thank Mr TN Kirway the Assistant Commissioner fur Farming Systems Research in the Ministry of Agriculture (Tanzania) fur facilitating this study and taking part in pre-testing the questionnaire Also we are thankful to Christian Mwakyembe Zakaria Malley Mathias Mkoma Francis Mosses Arnan Hallinga and Alatwisila Sakalani fur administering the questionnaire We are grateful to breeders and agronomists in the Maize Improvement Program at Uyole fur their technical input
We acknowledge the contribution of furmers in the study area fur responding to our questions willingly The contribution of the district and village level extension workers in all the three districts with regard to this report is highly appreciated
Last but not least we are greatly indebted to the secretaries Yohana Lukindo Issabella Minga and Aklilewerk Bekele fur expediting the publication of this report
VI
SUMMARY
Background to the study
This study was undertaken in order to document the level and rate of adoption of improved maize technologies in Mbeya region in the Southern Highlands of Tanzania The study had the following objectives
bull to investigate the rate and sequence of adoption bull to examine some characteristics of adopting and non-adopting fumers bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
Methodology
Initially literature review was made in order to get acquainted with research and extension services in the study area Based on literature review a reconnaissance survey was conducted in order to get first-hand infunnation about the study area Thereafter a questionnaire was administered to 100 funners in Mbeya Rungwe and Mbozi districts Fanners were either purposively or randomly selected depending on the required infunnation Multiple visits were made to each sample funner after each funn operation to capture their actual practices
Description of the study area
The study villages were from four different agro-ecological zones which practice dry season planting These included the Central Mbeya zone (1200 m) Mbozi Isuto and Ndola plateau and Poroto mountains (2400 m) and Tukuyu volcano and Rungwe valleys (2400 m)
Maize research in the Southern Highlands
Maize research in Southern Highlands started during the 1970171 cropping season at Uyole Agricultural Center (now MARTI Uyole) in Mbeya Municipality under the Tanzania Nordic Agricultural Project These activities increased steadily after 1974 following the establishment of the National Maize Research Program (NMRP) In 1981 maize research at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) The MIP was charged to develop appropriate varieties to improve parental lines of currently used commercial hybrids and to research on agronomic practices pest and disease control Recommended varieties in the study area include H614 UCA Kilima TMV-l and TMV-2
vii
Socio-economic characteristics of farmers
The majority (90) of male-headed households had almost 5 persons and about 42 of them had only 2 persons The average number of adult household members for female-headed fumilies was 24 persons Household size varied more in femaleshyheaded households than in the male ones The mean household size for male-headed families was significantly different from that of female-headed households More adults from male-headed fumilies worked on farms pennanently than the number of adults from female-headed households Most male family heads (96 ) had completed elementary fonnal education (below 7 years) while most female family heads (82 ) never went to school
Land tenure
Land ownership is communal and fanners have individual user rights Most male heads (825 ) and female heads (973 ) occupied a fann size of less than 10 acres Maize beans coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively Maize was the most important crop as it was cultivated by all respondents Seventy percent of male- and female-headed households did not tallow land Those who fallowed did so to increase soil fertility
Farm mechanization
Sixty-seven percent of all respondents used hand hoes and only 11 used oxen for cultivation and 22 used both hand hoes and oxen Most female fumily heads tended to use hand hoes (none used oxen) About balf of the surveyed households used hired labor mainly for land preparation wbile 21 hired labor for weeding and harvesting In both cases more female-beaded households hired labor than the male-headed households
Adoption of recommendations
Most fanners (70 ) started land preparation between July and November depending on the planting season i e dry or wet season Hovever where dry season planting was practiced plowing and planting were done concurrently About 57 of maleshyheaded households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
Planting practices
Almost all fanners adopted row planting Most female-headed households adopted row planting between 1986 and 1990 while male-headed households adopted before this period None of the sampled fanners follow the recommended levels of spacing but the majority who used 30 x 90 cm were very close to the recommended spacing
viii
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
ACKNOWLEDGEMENTS
The authors of this report thank the Commissioner of Research and Training Ministry of Agriculture Thnzania and CIMMYT fur providing financial and institutional support
We thank Mr TN Kirway the Assistant Commissioner fur Farming Systems Research in the Ministry of Agriculture (Tanzania) fur facilitating this study and taking part in pre-testing the questionnaire Also we are thankful to Christian Mwakyembe Zakaria Malley Mathias Mkoma Francis Mosses Arnan Hallinga and Alatwisila Sakalani fur administering the questionnaire We are grateful to breeders and agronomists in the Maize Improvement Program at Uyole fur their technical input
We acknowledge the contribution of furmers in the study area fur responding to our questions willingly The contribution of the district and village level extension workers in all the three districts with regard to this report is highly appreciated
Last but not least we are greatly indebted to the secretaries Yohana Lukindo Issabella Minga and Aklilewerk Bekele fur expediting the publication of this report
VI
SUMMARY
Background to the study
This study was undertaken in order to document the level and rate of adoption of improved maize technologies in Mbeya region in the Southern Highlands of Tanzania The study had the following objectives
bull to investigate the rate and sequence of adoption bull to examine some characteristics of adopting and non-adopting fumers bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
Methodology
Initially literature review was made in order to get acquainted with research and extension services in the study area Based on literature review a reconnaissance survey was conducted in order to get first-hand infunnation about the study area Thereafter a questionnaire was administered to 100 funners in Mbeya Rungwe and Mbozi districts Fanners were either purposively or randomly selected depending on the required infunnation Multiple visits were made to each sample funner after each funn operation to capture their actual practices
Description of the study area
The study villages were from four different agro-ecological zones which practice dry season planting These included the Central Mbeya zone (1200 m) Mbozi Isuto and Ndola plateau and Poroto mountains (2400 m) and Tukuyu volcano and Rungwe valleys (2400 m)
Maize research in the Southern Highlands
Maize research in Southern Highlands started during the 1970171 cropping season at Uyole Agricultural Center (now MARTI Uyole) in Mbeya Municipality under the Tanzania Nordic Agricultural Project These activities increased steadily after 1974 following the establishment of the National Maize Research Program (NMRP) In 1981 maize research at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) The MIP was charged to develop appropriate varieties to improve parental lines of currently used commercial hybrids and to research on agronomic practices pest and disease control Recommended varieties in the study area include H614 UCA Kilima TMV-l and TMV-2
vii
Socio-economic characteristics of farmers
The majority (90) of male-headed households had almost 5 persons and about 42 of them had only 2 persons The average number of adult household members for female-headed fumilies was 24 persons Household size varied more in femaleshyheaded households than in the male ones The mean household size for male-headed families was significantly different from that of female-headed households More adults from male-headed fumilies worked on farms pennanently than the number of adults from female-headed households Most male family heads (96 ) had completed elementary fonnal education (below 7 years) while most female family heads (82 ) never went to school
Land tenure
Land ownership is communal and fanners have individual user rights Most male heads (825 ) and female heads (973 ) occupied a fann size of less than 10 acres Maize beans coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively Maize was the most important crop as it was cultivated by all respondents Seventy percent of male- and female-headed households did not tallow land Those who fallowed did so to increase soil fertility
Farm mechanization
Sixty-seven percent of all respondents used hand hoes and only 11 used oxen for cultivation and 22 used both hand hoes and oxen Most female fumily heads tended to use hand hoes (none used oxen) About balf of the surveyed households used hired labor mainly for land preparation wbile 21 hired labor for weeding and harvesting In both cases more female-beaded households hired labor than the male-headed households
Adoption of recommendations
Most fanners (70 ) started land preparation between July and November depending on the planting season i e dry or wet season Hovever where dry season planting was practiced plowing and planting were done concurrently About 57 of maleshyheaded households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
Planting practices
Almost all fanners adopted row planting Most female-headed households adopted row planting between 1986 and 1990 while male-headed households adopted before this period None of the sampled fanners follow the recommended levels of spacing but the majority who used 30 x 90 cm were very close to the recommended spacing
viii
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
SUMMARY
Background to the study
This study was undertaken in order to document the level and rate of adoption of improved maize technologies in Mbeya region in the Southern Highlands of Tanzania The study had the following objectives
bull to investigate the rate and sequence of adoption bull to examine some characteristics of adopting and non-adopting fumers bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
Methodology
Initially literature review was made in order to get acquainted with research and extension services in the study area Based on literature review a reconnaissance survey was conducted in order to get first-hand infunnation about the study area Thereafter a questionnaire was administered to 100 funners in Mbeya Rungwe and Mbozi districts Fanners were either purposively or randomly selected depending on the required infunnation Multiple visits were made to each sample funner after each funn operation to capture their actual practices
Description of the study area
The study villages were from four different agro-ecological zones which practice dry season planting These included the Central Mbeya zone (1200 m) Mbozi Isuto and Ndola plateau and Poroto mountains (2400 m) and Tukuyu volcano and Rungwe valleys (2400 m)
Maize research in the Southern Highlands
Maize research in Southern Highlands started during the 1970171 cropping season at Uyole Agricultural Center (now MARTI Uyole) in Mbeya Municipality under the Tanzania Nordic Agricultural Project These activities increased steadily after 1974 following the establishment of the National Maize Research Program (NMRP) In 1981 maize research at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) The MIP was charged to develop appropriate varieties to improve parental lines of currently used commercial hybrids and to research on agronomic practices pest and disease control Recommended varieties in the study area include H614 UCA Kilima TMV-l and TMV-2
vii
Socio-economic characteristics of farmers
The majority (90) of male-headed households had almost 5 persons and about 42 of them had only 2 persons The average number of adult household members for female-headed fumilies was 24 persons Household size varied more in femaleshyheaded households than in the male ones The mean household size for male-headed families was significantly different from that of female-headed households More adults from male-headed fumilies worked on farms pennanently than the number of adults from female-headed households Most male family heads (96 ) had completed elementary fonnal education (below 7 years) while most female family heads (82 ) never went to school
Land tenure
Land ownership is communal and fanners have individual user rights Most male heads (825 ) and female heads (973 ) occupied a fann size of less than 10 acres Maize beans coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively Maize was the most important crop as it was cultivated by all respondents Seventy percent of male- and female-headed households did not tallow land Those who fallowed did so to increase soil fertility
Farm mechanization
Sixty-seven percent of all respondents used hand hoes and only 11 used oxen for cultivation and 22 used both hand hoes and oxen Most female fumily heads tended to use hand hoes (none used oxen) About balf of the surveyed households used hired labor mainly for land preparation wbile 21 hired labor for weeding and harvesting In both cases more female-beaded households hired labor than the male-headed households
Adoption of recommendations
Most fanners (70 ) started land preparation between July and November depending on the planting season i e dry or wet season Hovever where dry season planting was practiced plowing and planting were done concurrently About 57 of maleshyheaded households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
Planting practices
Almost all fanners adopted row planting Most female-headed households adopted row planting between 1986 and 1990 while male-headed households adopted before this period None of the sampled fanners follow the recommended levels of spacing but the majority who used 30 x 90 cm were very close to the recommended spacing
viii
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Socio-economic characteristics of farmers
The majority (90) of male-headed households had almost 5 persons and about 42 of them had only 2 persons The average number of adult household members for female-headed fumilies was 24 persons Household size varied more in femaleshyheaded households than in the male ones The mean household size for male-headed families was significantly different from that of female-headed households More adults from male-headed fumilies worked on farms pennanently than the number of adults from female-headed households Most male family heads (96 ) had completed elementary fonnal education (below 7 years) while most female family heads (82 ) never went to school
Land tenure
Land ownership is communal and fanners have individual user rights Most male heads (825 ) and female heads (973 ) occupied a fann size of less than 10 acres Maize beans coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively Maize was the most important crop as it was cultivated by all respondents Seventy percent of male- and female-headed households did not tallow land Those who fallowed did so to increase soil fertility
Farm mechanization
Sixty-seven percent of all respondents used hand hoes and only 11 used oxen for cultivation and 22 used both hand hoes and oxen Most female fumily heads tended to use hand hoes (none used oxen) About balf of the surveyed households used hired labor mainly for land preparation wbile 21 hired labor for weeding and harvesting In both cases more female-beaded households hired labor than the male-headed households
Adoption of recommendations
Most fanners (70 ) started land preparation between July and November depending on the planting season i e dry or wet season Hovever where dry season planting was practiced plowing and planting were done concurrently About 57 of maleshyheaded households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
Planting practices
Almost all fanners adopted row planting Most female-headed households adopted row planting between 1986 and 1990 while male-headed households adopted before this period None of the sampled fanners follow the recommended levels of spacing but the majority who used 30 x 90 cm were very close to the recommended spacing
viii
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Improved varieties
Of the sampled fanners 76 responded as having grown improved varieties at one time or other During the survey season 24 male-headed households and 74 female-headed households grew H614 and the local variety respectively Most fanners (53 ) adopted improved varieties between 1986 and 1990 However those who had adopted improved varieties (64 and 74 of male- and female-headed households) had discarded improved varieties at least once between 1986 and 1993 largely due to unavailability of seeds high cost bad taste and poor milling quality
Weeding
All respondents weeded their maize plots at least once during the crop cycle Sixty-five percent and 57 of male- and female-headed households weeded twice while 87 and 95 of male- and female-headed households used herbicides and none used oxen-drawn weeding cultivators
Soil fertility management
The use of inorganic fertilizers was highly associated with gender Seventy-eight percent and 46 of female- and male-headed households had used inorganic fertilizers Lack of money fertile soils and unavailability of fertilizers were the major reasons for not using fertilizers The majority of male-headed households used organic fertilizers than the female-headed households Most of the respondents (86 ) did not use fannyard manure The majority of the fanners adopted fertilizer use between 1976 and 1990
Pests and diseases control
Stalk borers rodents birds and maize streak virus (MSV) were major pests and disease Eighty-one and 83 of male- and female-headed households indicated that their local maize cultivar was the most affected by pests and diseases Sevenshyty-seven percent and 86 male- and female-headed households did not do anything to prevent their crop against pest and disease damages
Harvesting and storage
Maize was harvested between April and June for dry season planters and between June and July for wet season planters After shelling maize was mainly stored in gunny bags (49 of the respondents) Twenty-tG percent and 24 of male- and female-headed households stored unshelled maize in traditional structure called Vihenge Seventy-one percent and 73 male-and female-headed households treated their maize with Actellic dust and 8 male-headed households used pyrethrum flowers
ix
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Factors affecting adoption
Cost of technology environmental stress lack andor timely availability of the technologies and the lack of infurmation on new technologies have influenced the adoption of maize production technologies by the sample furmers
Technologies which require little cash outlay such as row planting were easily taken up by furmers Row planting was adopted by 96 of sample furmers
EINironmental stress constrained adoption and some of the recommendations specially where maize is planted during the dry season Farmers who dry planted their maize did not apply basal fertilizer Others did not perform second weeding Prevalence of pests and diseases specially stalk borer and MSV might have affected adoption of improved maize varieties
Lack of andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent and 75 of male- and female-headed households cited unavailability of fertilizer as the reason why they were not using it Fifty-fi-e percent and 66 of male- and female-headed households had discarded H614 because of unavailability of seed
Half of the furmers in the survey area (46 and 57 of male-and female-headed households respectively) had never been visited by extension workers Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seed row planting fertilizer use and use of herbicides In respect to extension contact both male and female furmers had received visits in almost equal proportion
Implications for research extension and policy
The problem of stalk borer needs to be addressed by developing maize varieties tolerant of this pest Research has addressed the problem of MSV by developing a variety known as TMV-I that was supposed to be resistant to this virus although furmers were not generally aware of this variety The range of hybrid maize varieties that are acceptable and affordable by tarmers needs to be widened At present it would appear that H614 is the most available and acceptable hybrid maize variety And even for H614 the disadoption rate is very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is a great need for extension service to improve its contact with furmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of knowledge on how to use herbicides for example constrains the adoption of this technology specially by male-headed households
x
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
while the lack of infunnation on such technology as TMV-I maize variety that is resistant to MSV might have limited its use This might also be a reflection of the lack of research-extension linkage Given the constraints fuced by the extension service in meeting the needs of individual funners extending messages to groups of funners rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on prices of seed and fertilizer due to structural adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed and fertilizer technologies and its critical role in increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seed and fertilizer subsidy in the short to medium tenns
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SG-2000 Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to lack of money to purchase it compared to 29 of male-headed households Lack andor timely availability of technology are major constraints to its adoption therefore policy should create a fuvorable environment for provisions of inputs specially seed and fertilizer by fucilitating the development of a viable private sector
In the short to medium tenns the provision of credit to traders to purchase seed and fertilizer and also construct storage fucilities will assist in promoting the developshyment of the private sector In the future general infrastructural development such as access roads and local markets will be needed specially to reduce the long distances travelled by funners to obtain these inputs Alternatives such as giving researchers at Uyole the mandate to produce seed and sell to funners should be explored
Policy-makers should examine the imbalance between male and female furmers in tenns of access to production resources such as land and capital as well as to infunnation on new technologies For instanGe male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no fonnal education More female-headed households fuced the problem of unavailability of seed and fertilizer (59 and 75 respectively) compared to maleshyheaded households (35 and 47 respectively) This certainly calls for a more concerted effort by policy to target female-headed households not only on equity grounds but also to assist in increasing maize production which will enhance the nationS food security
Xl
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Introduction
Maize is the single most important food crop in Tanzania Over 80 of the population depends on it It accounts for more than 40 of the total cultivated area Per capita conSumption is around 113 kg per year and maize contributes 61 of the total calories in the diets of Tanzanians (Mayagilla 1988) Maize is produced in all 20 regions of the main land but there is surplus production in the Southern Highlands ie Iringa Mbeya Ruvuma and RukWcl all of which account for about 46 of the national production (Moshi and Nnko 1989) Furthermore the Southern Highlands account for nearly 90 of the total maize purchased for the National Food Security Granary (Mussei and Shiyumbi 1992) The area planted with maize has been increasing through the years especially after the 1970s (Croon et aZ 1984) For instance between 1980 and 1988 the acreage under maize increased from 1529000 to 1867000 ha as a result production rose from 17 to 31 million tones (Moshi and Nnko 1989)
Over 80 of maize production in Tanzania is produced by small-scale farmers under a wide range of management practices climatic conditions and socioeconomic circumstances The rest is produced by village farms private and public large farms There are three major maize recommendation zones First the highlands zone with altitudes of above 1500 m and growing period of 6-8 months Second the intermediate zone 900-1500 m This zone is divided into wet and dry sub-zones The wet sub-zone gets over 1100 mm rainfall with 4-5 months of growing period The dry sub-zone gets less than 1100 mm rainfall with a 3-4 months growing period Third the lowland zone with altitudes of 0-900 m and 3-4 months growing season (Semuguruka 1988)
Maize production in the Southern Highlands started early this century mainly in Iringa region and later expanded to other regions specially Mbeya and Ruvuma in the 1950s and into Rukwa region in the 1970s (Mussei and Shiyumbi 1992) Maize has replaced traditional small cereals and the zone is generally referred to as the big four in terms of maize production
The Problem
The maize improvement program at Uyole Agricultural Research and Training Institute (MAKfI-Uyolej (formerly Uyole Agricultural Center) which started during the 1970171 cropping season has contributed greatly to the success of maize production in the Southern Highlands Working in collaboration with national and international research centers the Maize Improvement Program has tested adapted and released several improved maize varieties and their accompanying agronomic
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
recommendations fur various agro-ecological zones of the Southern Highlands However although these recommendations are currently in use the extent to which they have been adopted by funners is not well known quantified and documented There is a wide gap between the funners average maize yields of about 15 tha and that of research of 72 tha
Objectives of the Study
This study was initiated to investigate and document adoption levels and to draw implications for research extension and policy The specific objectives were
bull to investigate the rate and sequence of adoption of recommendations bull to examine the characteristics of adopting and non-adopting funners bull to examine the characteristics of technologies that promote adoption and bull to draw implications for research extension and policy
2
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Metll0dology
The methodology used in this study involved reconnaissance survey key infi)lmant interviews review of secondary sources questionnaire design pre-testing and implementation
Reconnaissance Survey
Based on infurmation from existing literature too different planting seasons ie dry and wet were identified A reconnaissance survey was thereafter undertaken in the two rones ie Umalila highlands and Utengule lowlands in Mbeya district where the two planting seasons are found The objectives of the reconnaissance survey were
bull to get acquainted with the study areas bull to get general understanding of actual furmers practices and constraints and bull to assist in the selection of representative sample villages
To identify and select the study areas infurmal interviews were conducted with furmers and extension staff at village district and regional levels as key infurmants
Questionnaire
The questionnaire was developed in collaboration with the Maize Improvement Program (MIP) agronomists on the basis of infurmation obtained from the reconnaissance survey It was later pre-tested by a team of scientists from Farming Systems Research (FSR) and the MIP with some participation by the Assistant Commissioner (FSR) and the CIMMYT regional economist The formal survey started in July 1992 for the dry season planting areas and November 1992 for the wet season planting areas and completed in September 1993
Sampling Procedure
Although MARfI-Uyole has the research mandate over the four regions of the Southern Highlands only Mbeya was selected for this study due to logistic reasons It is anticipated that this kind of study will be extended to other region of the zone in the future
Purposive and multi-stage random sampling procedures were used to select study districts wards villages and furmers With the assistance of the regional and district extension staff major maize producing districts and divisions were purposively
3
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
selected namely Mbozi for wet planting season Rungwe for dry season planting and Mbeya for both wet and dry season planting
In districts with wet planting seasons sample divisions were randomly selected whereas in districts with dry season planting regimes sample divisions were purposively selected because it was not everywhere furmers practice dry season planting In each study division regardless of the planting season too wards and one village were randomly selected The criteria for selecting furmers were random for all male-headed households but for female-headed households it was purposive because such households were not common in the study area The reason to include female-headed households was to see if there are gender differences in technology adoption
Based on the importance of each district in terms of maize production 40 furmers were selected each from Mbeya and Mbozi districts while the remaining 20 were selected from Rungwe district The reason behind this was that relatively less maize was produced in Rungwe district than Mbeya and Mbozi districts Out of the 100 sample farmers 37 were females from female- headed households (Table 1) Although it seems that most female respondents lived in rural Mbeya the chi-square test showed non-significant result (x2=315 df2) The indication is that gender and districts are not associated Therefore it can be said that male- and female-headed households have been selected from each district with similar proportion
Table 1 Household head by gender and district
District
Gender Rural Mbeya Rungwe Mbozi Total
Male 24 10 29 63
Female 16 10 11 37
Total 40 20 40 100
4
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Description of the Study Area
The location of Mbeya region relative to other regions of the Southern Highlands is shown in Figure 1 while the location of the study area in Mbeya region is shown in Figure 2
A large proportion of the Southern Highlands has altitudes over 900 m and fairly unifurmly distributed unimodal rains that frequently exceed the minimum requirement for maize production Temperatures are near optimum throughout the growing season usually above 15degC Water-holding properties and pH of the soils are also good (Marandu et al 1989) According to Croon et al (1984) the study areas full into the following agro-ecological wnes
bull The U songweU tengule wards in Mbeya Rural district full under the Central Mbeya wne VIII In this area which rises up to 1200 m the average raiufull is about 700 mm Maize and beans are staple crops in the area Other enterprises are coffee banana cattle goats sheep pigs and chicken In this wne all crops are grown during the wet season
bull Ilembo and Santilya wards also in Mbeya Rural district full under wne IX called the Ilembo plateau and Poroto mountains The altitude ranges between 1900 and 2400 m with the annual rainfull ranging between 2500 and 3500 mm The growing season is nine months The soils are deep dark loam with bulk density and high to moderate fertility Some parts of the wne have deep dark sandy loams and loamy sands of moderate to low fertility (Nalitolela 1990) Maize beans pyrethrum potatoes peas finger millet and wheat are major crops in this wne Farmers also keep cattle goats sheep pigs and chicken Although all seasons are suitable for growing crops maize is planted during the dry season
bull Kandete and Mpombo wards full under the Isoko dissected plains fukuyu volcano and Rungwe valleys wne XI These are hilly and dissected lava plateaus with altitude range of 900 to 1800 m Raiufull ranges between 1900 and 2400 mm with some precipitation every month In addition to maize other crops grown in the wne are tea banana coffee beans peas cocoyams and assorted fruits Livestock types include cattle sheep goats pigs and chicken These areas are fumous for keeping improved dairy cattle
bull Vwawa and Halungu wards are found in wne IV known as Mbozi Isuto and Ndo1a The altitude ranges between 1900 and 1800 m while the annual
5
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
rainfull is between 1100 and 1200 IllIll The growing season is about 65 months which is adequate for a good maize production Dominant soils are brownish ash red and moderately fertile This zone produces surplus maize in the region Other crops which are produced in the zone include beans coffee cassava sweet potatoes finger millet and banana
LEGEND middot _INTERNATI~L
BOJNOARf _ _ REGIaIAl BOUNDARY
BOOtES
TANZANIA
ZAIRE
ZAMBIA
Figure 1 Tanzania the location of the four region of the Southern Highlands zone
6
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
LEGEND ___ INTERNATIONtltl BOO~-l
RUKWA --_- --- REGIONAl BOUNDARY -_- WATER BODIES
(~ STUDY AREA
i
MBEYA
J
I -----
t I
Figure 2 Mbeya Region the location of the study area
7
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Maize Research in the Southern Highlands
Maize research in Tanzania started during the colonial period From 1940s there were research activities concentrated mainly on varieties selection fertilizer application rates and materials plant densities and time cf planting During that period research activities were conducted independently at each research institute During the 196667 cropping season a coordinated maize research program was initiated charged with the responsibility to undertake yield trials on selected sites throughout the country (Semuguruka 1988)
In 1974 under the technical leadership of CIMMYT the National Maize Research Program (NMRP) was established at Ilonga Agricultuml Research Institute serving as the coordinating center The program was responsible fur all phases of maize research from varietal development and management practices to verification on farmers fields in different agro-ecological zones (Marandu et al 1989 Lyimo and Temu 1992)
In the Southern Highlands maize research started during the 1970171 cropping season at Uyole Agricultural Center (now MAlITI-Uyole) under the TanzaniashyNORDIC Agricultural Project Following the initiation of the NMRP in 1974 research activities at Uyole increased tremendously especially in the area of village trials which facilitated the verification of agronomic packages under fanners field conditions while at the same time serving as demonstration plots (Lyimo and Temu 1992)
In 1980 maize research activities at Uyole expanded under the Southern Highlands Maize Improvement Program (MIP) Essentially the MIP had four broad objectives (Marandu et al 1989)
bull to develop varieties suitable for the environment and fanning conditions of the highlands
bull to maintain and improve parental lines of all current conmlercial hybrids used in the country
bull to do research on agronomic practices suitable for the agro-ecological zones and funning systems in the Southern Highlands and
bull to monitor pests and diseases and recommend control measures
Table 2 shows the characteristics of some commercial varieties some of which are results of the combined efforts between Uyole and other national and international
8
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
research institutions and adapted fur the Southern Highlands
Table 2 Characteristics of released improved maize varieties
Variety Year Yield Altitude Maturity released potential Period
(t ha1 ) (days)
H614middot 1977 and 1979 70 high 180-200
H6302 1976 80 high 180
H632 1965 60 intermediatehigh 170-180
UCA 1966 45 intermediate 140
Kilima 1983 48 intermediate 140
Kito 1983 35 low 90-100
TMV-1 + 1987 40 lowintermediate 130
TMV-2 1987 70 high 170
Source Personal communication with the maize breeder at MARTI Uyole was released twice in 1977 using EC 573 C2 as the male parent in
the final cross and in 1979 using EC 573 C5 as the male parent + tolerant to MSV
Recommended Maize Technology
arieties
Table 3 Recommended maize varieties for different wards and agro-ecological zones
Ward Agro-ecological Recommended Variety ~one
KandeteMpombo X H6302 H614 UCA Kilima
lIemboSantilya IX H614 H6302 TMV-1 TMV-2
UtenguleUsongwe VIII H614 Kilima TMV-l TMV-2 UCA
VwawaHalungu IV H6302 H614 H632 UCA Kilima TMV-2
9
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Planting
Infurmation on time of planting fur wet season was available while that for dry season planting was not For the wet seasons it is recommended to plant maize immediately after the commencement of rains which is normally after 15 November in most parts of the Southern Highlands and not later than 15 December However for the past fuur years the rainfall pattern has shifted thus affecting the time of planting Rains start during the second half of December and this is when farmers should start planting It is estimated that timely planting and weeding should raise yields from 700 to 1200 kgha (Lyimo and Temu 1992) Studies in Kenya have shown that if planting is delayed for tM) days 70 kg of grainlhaday would be lost (Cooper and Law 1976) In the Southern Highlands of Tanzania the grain yield loss due to the late planting of recommended hybrids was 62 kgha for a days delay after the first rains (Lyimo and Temu 1992)
Planting in rows is recommended in order to achieve the recommended plant population and to facilitate weeding Under fertile soils optimum plant popUlation is about 45000 plantsha whereas under low soil fertility the population ranges between 22000 and 33000 plantsha Plant population is also determined by spacing and seeding rate The recommended seeding rates are indicated in Table 4
Fertilizer
In the Southern Highlands nitrogen and phosphorus are the major limiting nutrients fur maize production Studies have shown that improved varieties require substantial quantities of mineral nutrients for their vegetative and grain development for instance a crop that produces 5-6 tha will have removed 100-150 kg of nitrogen and 40-60 kg of P20 S per ha from the soils (Prasad 1978) The use of both inorganic and organic fertilizers can lead to high yield Research undertaken at Uyole showed that when low doses of Nand P (ie) 40 and 15 kgha respectively) supplemented with 20 tha of farmyard manure (FYM) the grain yield was 71 tha compared with 4 03 tha when the same rates of Nand P were used alone and 512 tlha when FYM at the rate of 20 tha was used alone (Lyimo and Temu 1992)
Fertilizer recommendations are based on soil types and nutrient deficiency For basal application TSP and NPK at the rate of 100-400 kgha are recommended For top-dressing CAN at the rate of 200-400-600 SA at the rate of 250-500-750 and urea at the rate of 100-200-350 kgha are recommended
10
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Table 4 Recommended spacing and seeding rate
Spacing Seeding rate Suitability (cm)
75 x 30 1-1-1-1 or 1-2-1-2 planter 75 x 60 2-2-2-2 or 2-3-2-3 hand hoe 75 x 90 3-3-3-3 or 3-4-3-4 hand hoe 90 x 25 1-1-1-1 or 1-2-1-2 planter 90 x 50 2-2-2-2 or 2-3-2-3 hand hoe
Source Temu (1991)
Weeding
Weeds seriousiy affect maize yield In the Southern Highlands of Tanzania yield reduction range between 50 and 100 of the potential yields (Croon et al 1984 UAC 199293) Similar trials at Ilonga in the Eastern wne recorded a yield loss of 25 when the first (these rates show minimum medium and maximum levels for different categories of farmers ie poor middle and rich farmers) weeding was delayed until four weeks after planting while in the Lake wne delaying weeding for six weeks resulted in 65 yield reduction (Matowo and Mjema 1988) Thus it is recommended to perform at least too weedings using either hand hoes cultivators or herbicides The first weeding is recommended 2-3 weeks after emergence while the second should be performed when the plants are at waist height (90-100 cm) (Tale 5) Crop rotation can also control some types of weeds especially striga
If farmers decide to use herbicides the following are recommended for preshyemergence application Attrazine Laddock Attrazine-Metalachlor Lasso-Attazine Alachlor and 2-4D and post-emergence application of Paraquate (Gramaxone)
Table 5 Effect of different weeding regimes on maize grain yield
Yield Increase over Weeding regimes (tlha) control ()
No weeding 228 100 Once at 10 cm height 217 183 Once at 30 cm stage height 388 170 Once at 50 cm stage height 4 09 179 Twice at 10 cm stage height 532 233 Twice at 30 cm and
70 cm height 541 237 Thrice at 10 50 and
90 cm height 542 238
Source Lyimo and Temu 1992
11
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
Pest and disease control
Major pests and diseases in the Southern Highlands are cut wonns stalk borer rodents birds and MSV Like weeds insect pests and diseases cause severe yield losses if they go unchecked In Malawi diseases have been recorded to cause up to 10 yield loss (Ngwira 1989) while in Kenya the yield loss was between 13 and 70 (Ochor et al 1989) In Zimbabwe MSV causes a yield loss of up to 43 (Mguni 1988) In Tanzania Mduruma et al (1988) observed that plants infected with MSV less than a week after gennination resulted in severe yield reduction Thus yield losses depend on the age of the plant at infection As for pests especiaU y stalk borer the damage may be as high as 20 (Lyimo and Temu 1992)
As a way of controlling stalk borer and cutwonn Thiodan-EC35 and Cypennethrin or Sumicombi are reconmlended A local herh called Utupa (Tephrosia vegilii) is also recorrunended Timely planting and crop rotation can reduce the damage caused by pests and diseases Scaring and trapping are also effective control measures for birds and rodent respectively
Research results at Uyole have shown that under good management grain yields of 6-7 tonsha can he achieved using appropriate maize cultivars Table 6 contains the sunm1ary of reconunended practices for maize production in the Southern Highlands while Table 7 shows the potential yield of high- and mid-altitude maize cultivars in different locations
Table 6 Recom mended practices for maize production
Management Yields estimates practices Characteristics (kgha)
Zero management Extremely low 300 One timely weeding 2-3 weeks after planting 700 Timely planting At optimum time (2 weeks 1200
after rains) Fertility improvement 20 kg P + 40-50 kg Nha 2100
(basal application) Optimum plant population 75 x 60 cm 2 plants per hill 2700 Improved seed HybridComposit e 3800 Further fertility
improvement 50 kg Nha + second weeding 6000 Pest control Control of stalk borer 7200
assumes a low soil fertility Source Lyimo and Temu (1992)
12
Table 7 Potential yield (tha) of high and mid-altitude maize cultivars in different locations in the Southern Highlands
High altitude ( gt 1500 m) Mid altitude (1000-1500 m)
10 years 5 years mean 199091 mean 199091
H6302 76 83 H632 3 9 37 H614 74 85 Kilima 46 42 TMV-2 67 9 5 TMV-1 4 5 54
UCA 42 53
Source Lyimo and Temu (1992)
13
Socio-Economic and Demographic Characteristics
of Farmers
Demographic Characteristics
Household size and labor availability
There were 297 adult fumily members above the age of 18 years in all households On average there were about 33 (SE=0 19) such adult fumily members in maleshyheaded fumilies which ranges from 7 to 2 persons The majority (90 ) of maleshyheaded households had at most 5 persons and about 42 of them had only 2 persoIlS On the other hand the average number of adult household members for female-headed fumilies was 248 adults (SE=027) Household size ranges from 9 to 1 Seventy percent of these fumilies have at most 2 adults and 84 at most 3 adults Household size tended to vary more in female-headed households than in the male ones
T-test and Analysis of Variance (AN OVA) showed that mean household size for male-headed fumilies is significantly different from that of female-headed household (t=81 df=98) This difference was not affected by districts ie differences between male-and female -headed fumilies were consistent and approximately equal in each district Mean household size for both male and female however was the same for all districts
The majority of adult household members (208 from male-and 76 from feshymale-headed fumilie s) work on farm permanently therefore an average of 33 (SE =017) male-headed furnily members work on farm permanently The maximum and the minimum male fumily members working on farm permanently coincided with the maximum and the minimwn for male-headed household size (7 and 2 respectively) Forty-six percent of these fumilies have 2 and 93 at most 5 members working on farm permanently Female-headed fumilies have an average of 2 (SE =023) members working on fann permanently Members working 011 faml
permanently range from 8 to I with 87 of these fumilies having at most 3 adults In this regard therefore female-headed fumilies showed larger range of variation than that of male-headed fumilies
More number of adults from male-headed furnilies work on farm pennanently than the number of adults from female-headed household The mean 33 (SE =017) for
14
male-headed fumilies is significantly greater than the female-headed fumilies (mean=208 SE=023) (t= 114) Nevertheless there was no difference between the districts and no interaction was observed between districts and gender
About 12 adults in male-headed fumilies (mean=019 SE=008) and 9 in female-headed fumilies (mean=024 SE=009) worked on the funn on part time basis The maximum number of such adults was 4 and 3 for male-and female-headed fumilies respectively The mean number of adults working part time in male-headed fumilies was not significantly different from the female-headed fun1ilies (t =0 5) The majority of households (90 male and 87 female-headed fumilies respectively) did not work on furm part time
It seems generally that there was no tradition of working off-furm permanently by farmers in the three districts under study as only 2 and 1 adults from female- and male-headed fumilies respectively were reported as working off-furm permanently
Household head age and settlement
Mean male household head age (mean=428 SE=207) was not significantly different from the female (mean=468 SE=189 t=19) There was however a relatively larger variation in the age of male than in female heads (68 and 45 for male and female respectively) About 88 of male fumily heads were at most 60 years old and only about 5 were older than 79 years while 83 female heads reached at most 60 years and only 17 were older than 60 years This indicates that male heads are composed of both young and old funners while most female heads are at a middle age This might be due to the tendency that females do not take the responsibility of heading the fumily at an early age such responsibilities occur when they are divorced separated or widowed usually at later ages
Irrespective of gender average years lived in the village was lower than average age of household head These values are significantly different from each other for male and female heads (mean=304 and SE=18 mean=396 and SE=295 for male and female respectively) Table 8 shows the proportion of the furmers who lived for certain years in the village The proportion of female heads was almost the same for various categories of ages lived in the village
Educational level
The level of education among furmers in the study area is generally low The majority of male heads (96 ) completed at most elementary grades (7 years or below) while about 5 of them completed or nearly completed secondary school (11 years or above) and only one male head had college education (14 years) In contrast most female heads (82 ) never went to school and only 11 completed
15
at most elementary school In addition to cultural factors which usually affect female education in Africa such low level of education can also be attributed to the fact that female heads are generally older than male heads The mean years of education for male heads (mean = 48 SE=OA) is significantly greater than that of female heads (mean = 09 SE=037)
Table 8 Range of years lived in the village by gender of household heads
Range of Household heads years lived Male Female
() ()
3 - 29 508 270 30 - 40 25 7 27 1 41 - 50 134 24 3 gt 50 110 216
Socio-Economic Characteristics
Land tenure
Like elsewhere in Tanzania in the study area land ownership is communal and fanners have individual user rights Land is inherited by the male fumily member or son in case of death or retirement of the household head
The surveyed households had occupied a total of 54636 acres of land of which about 88 had been cultivated (Table 9) This means that a considerable size of land was not being cultivated for various reasons
Table 9 Farm size and cultivated area by gender of household head
Farm size Cultivated area (acres) (acres)
Statistics Male Female Male Female
Mean 647 375 580 319 SE (mean) 053 0 38 048 028 Maximum 2400 1200 2400 900 Minimum 050 130 050 100
16
However when disaggregated by gender mean num size for male-and feshymale-headed fumilies is not significantly greater than their respective cultivated areas (t=09 t=12 respectively)
Most male-(825 ) and female-headed poundunilies (973 ) occupied a furm size of less than 10 acres All female-headed fumilies cultivated below 10 acres although 27 of them owned a furm size of up to 12 acres Male-headed fumilies on the other hand cultivated up to 24 acres of land They showed higher variability both in their holdings of total furm size and cultivated area than female-headed poundunilies The range for the two group was considerably different (Table 9) This might be attributed to the fuct that the age of male heads is highly and positively associated to furm size (regression of furm size on age is significant) Therefore there is a tendency that young male heads own small furm size and the furm size increases with age
Mean furm size and cultivated area for male heads was significantly different from that of female (t=42 and 47 respectively) However this difference does not interact with (or depend on) differences in districts In addition no considerable difference was observed between districts It seems that female~headed households used their holding more efficiently for cultivation than male heads in that most of them (97 ) had furm size of at most 10 acres of which all of it was cultivated This could be because male heads either have large land area all of which they could not cultivate or they used uncultivated area for livestock keeping as male-headed households owned more number of livestock than female heads
Based on area cultivdted maize was the most popular crop in the area followed by bean and coffee Maize bean coffee and other food crops occupied 56 15 12 and 5 of the total cultivated land respectively (Table 10) All furmers planted maize during the survey season But most furmers either did not plant other crops or planted them on less than one acre except beans which were cultivated by some funners on a larger land area The majority of male heads planted between I and 5 acres of maize (87 ) up to 2 acres of beans (94 ) and up to 1 acre of other crops Most female-headed fumilies on the other hand planted at most 3 acres of maize and 1 acre of all other crops Mussei and Shiyumbi (1992) found that on the average area under maize in Mbeya district was 275 acres while that of beans coffee potato and finger millet was 05 025 025 and 0125 acres respectively Area for bean potato pyrethrum and coffee showed much variation among male-headed fumilies Some of them planted these crops on large areas and others on small areas For female-headed fumilies however it seems that areas under these crops were much more uniform (Thble 10) Mean area under most of the crops for male-headed fumilies was significantly greater than that of female-headed fumilies
Farmers in the three districts had grown maize in at most 4 plots of varying sizes in the survey year On average male heads planted on 1 98 (SE=014) plots and
17
female on 181 (SE=016) plots These values were not significantly different from each other hence both male and female-headed families planted maize on equal number of plots in the survey year In most cases maize plots for male heads were found to be larger than that of female heads This indicates that female household heads tended to grow maize on smaller plots About half of the fimners grew maize on a single plot of land (48 and 51 for male and female households respectiveshyly) (Table 11) Greater proportion of female-headed households planted maize on 3 plots than male-headed households while the number of male-headed households planting maize on 4 plots of land was greater than the number of female-headed households planting maize on the same number of plots
On practice of fallowing land it was found that about equal proportion (70 each) of male-and female-headed households did not fallow their maize land An interesting point regarding gender however was that although female heads have smaller farm when compared to male heads they fullow land in equal proportion to male-headed households This might be womens strategy to conserve soil fertility due to their inability to purchase inorganic fertilizers
Farmers fallowed land mainly to increase land fertility and reduce labor pressure Eighty seven percent of those fullowing land did so to increase land fertility while the remaining 13 fullowed in order to release labor pressure for the next plowing season There was no difference between genders regarding reasons for fullowing (Table 12) In general therefore only 26 of all farmers fullowed their land
Farm mechanization
Hand hoes constituted the major source of power among the fimn families as reported by 67 of all respondents Only 11 farmers (11 ) all from Mbozi district used oxen only for cultivation Twenty-tQ fanners used both oxen and hand hoe for cultivation and half of these were from Mbozi Among the 33 oxen users 20 farmers owned a pair of oxen while the rest hired from neighbors These results are similar to the findings of the ASSP (1992) baseline study
From tabulations and chi-square tests it was found that gender and source of farm power were associated (i=5 3 df=2) Femaleheaded households tended to use hand hoe but not oxen In contrast 26 of male-headed households used oxen or both hand hoe and oxen The reasons for this difference might be because feshymale-headed households have smaller land area than the male ones hence did not require oxen or female-headed households cannot afford to buy or hire oxen In about half of the sample households fumily labor alone is not enough to work on maize farm hence labor was hired for land preparation weeding and harvesting Male farmers tend to hire more labor for land preparation weeding and harvesting than female-headed families (Table 13) This might be because male-headed households owned large maize areas compared to female-headed households
18
Table 10 Area under various crops (acres) by gender of household head
Statistics
Crop Gender Mean
Maize Male 311 Female 199
Bean Male 089 Female 039
Potato Male 005 Female 0 01
Other food Male 027 crops Female 025
Pyrethrum Male 035 Female 009
Coffee Male 0 79 Female 0 20
Tea Male 016 Female 018
Other cash Male 0 20 crops Female 007
SE (mean) Max
024 023
90 80
013 005
60 10
0 028 0009
10 03
006 006
20 10
012 005
60 10
021 005
120 10
007 009
30 25
009 005
50 15
Table 11 Number of maize plots by gender of household head
Number of plots Freguencll (l maize grown Male Female Total
1 476 514 49 2 206 216 21 3 175 21 6 19 gt 4 143 54 11
19
Table 12 Reasons for fallowing land by gender of household head
Reasons
Gender Land fertility Labor pressure Ilot fallow Total
Male 17 (27) 2 (3) 44 (70) 63 Female 9 (24) 2 (6) 26 (70) 37 Total 26 (26) 4 (4) 70 (70) 100
1 Percentages in brackets
Table 13 Hired-labor used for maize farming by gender of household head
Land QreQaration Weeding Harvesting Gender Yes No Yes No Yes No
Male 36 (57) 27 (43) 16 (25) 47 (75) 20 (32) 43 (68) Female 15 (41) 22 (59) 5 (14) 32 (88) 5 (14) 32 (86) Total 51 (51) 49 (49) 21 (21) 79 (79) 25 (25) 75 (75)
Livestock
Livestock are kept in all districts of the area The main types of livestock included cattle goats sheep and poultry The majority of the farmers (84 ) kept livestock It was also observed that the proportion of female-headed households keeping livestock is very similar to that of male heads (81 and 86 respectively) The Chishysquare test therefore showed the absence of association between gender and livestock keeping However when the data were disaggregated into cattle and other livestock a different pattern emerged About 76 of female heads did not keep cattle while about 21 kept at most 5 cattle and the remaining 4 kept more than 5 cattle On the other hand about 46 male-headed fumilies did not keep cattle while 46 kept at most 5 cattle and about 6 kept between 6 and 10 and the remaining 2 kept more than 10 cattle Chi-square tests indicated that there is a significant association (~ = 189 df= 1) between gender and cattle ownership cattle keeping being more associated with male-headed fumilies while other livestock keeping does not differentiate between gender (i = 030 df= 1) They were kept in almost equal proportion by both groups
Mean cattle and other livestock kept by male farmers was significantly greater than that of female-headed fumilies (Table 14) Although about equal proportion of both male and female farmers kept other livestock the number kept by individual male farmers tended to be greater than that of female farmers
20
Table 14 Mean number of cattle and other livestock kept by gender of household heads
Statistics
Livestock Gender Mean SE (mean) Max T-value
Cattle Male 209 0 31 1200 Female 078 027 600 318 Total 161 023 1200
Other Male 933 106 3900 livestock Female 490 0 90 2300 318
Total 771 078 39 00
The mean number of other livestock kept by both male and remale furmers was greater than the mean number for cattle The fact that poultry sheep goats etc were counted as they were without being converted into livestock unit must have exaggerated mean number for other livestock
In order to estimate the extent of relation of various crop areas to livestock a correlation analysis was made It was found that there is a moderate relationship between number of livestock and areas under some of the crops Number of cattle and other livestock showed a positive relation with food crops and no relation with cash crops Since by-products of food crop production such as weed stover forage etc can be used for reeding livestock there is a possibility that the large number of livestock is associated with large area of food crop
21
Adoption of Recommendations
Land Preparation and Planting
Land preparation
Most of the respondents (70 ) started land preparation during the first week of the months of July-NltNember (Table 15) The time and duration of land preparation varied according to the number of plots one has labor availability and the season of planting i e dry or wet season Where dry season planting is practiced for instance land preparation goes hand in hand with planting between May and September
Planting maize
Seventy-nine percent of all farmers responded as having grown maize in the same plot consecutively Average number of consecutive seasons in which maize was grown on the same plot was 435 (48 for male and 36 for female) In case of maize about 20 of the total respondents (30) fallowed their maize field compared to about 30 of all farmers that fallow land
The practice of fallowing maize and growing it in the same plot did not differ between male and female farmers Both maize fallowing (t= 15 df= 1) and growing maize in the same plot (t=005 df= 1) were not associated with gender ie both male and female-headed households responded in the same proportion
Time of planting maize was also assessed to see the ideal planting period for maize varieties Accordingly 63 of male-headed households responded as planting maize during July-August and 100 female-headed households during NovembershyDecember in their first plot Generally both female-and male-headed households tended to plant the other maize plots at the same time as the first plot About 57 of male-headed households planted maize at the recommended time while almost none of the female-headed households followed this recommendation
22
1
Table 15 Time of land preparation by plot of land and gender of household head
Plots
Period Gender 2 3 4
Male 11 4 1 1 May-June Female 8 2 0 0
Sub-total 19 6
Male 24 15 15 8 July-Sept Female 16 14 9 2
Sub-total 40 29 24 10
Male 20 8 3 0 Oct-Nov Female 10 2 1 0
Sub-total 30 10 4 0
Male 8 6 2 1 Dec-Jan Female 3 0 0 0
Sub-total 11 6 2
Male 0 30 42 53 Not Female 0 19 27 35 preparing Sub-total 0 49 6 88
The time refers to the first week of the month
Planting practices
Almost all turners adopted row planting (Table 16) Chi-square test showed that there was a significant association between types of row planting and gender (-=97 df=2) Most female furmers tended to use row spacing of 30 x 90 cm and more proportion of male-headed households tended to use 60 x 90 cm of row spacing for maize planting
Row planting was adopted as early as 1970 (Table 17) This might be because of the fumous political campaigns popularly known as Kilimo cha kufu na kupona (Agriculture for death or survival) following the acute nationwide food shortages during the first half of the 1970s
23
Table 16 Spacing between rows by gender of household head
Spacing (cm) Male Female Total
30 x 60 13 (21) 8 (23) 21 (22) 30 x 90 30 (49) 23 (66) 53 (55) 60 x 90 18 (30) 4 (11) 22 (23) Total 61 (63) 35 (37) 96 (100)
Table 17 Row planting by year of adoption and gender of household head
Period adopted Male Female Total
1970-75 15 (25) 3 (9) 18 (19) 1976-80 21 (34) 8 (22) 29 (30) 1981-85 8 (13) 9 (26) 17 (18) 1986-90 17 (28) 15 (43) 32 (33) Total 61 (63) 35 (37) 96 (100)
When comparing the funners in the pattern of adopting row planting (Thble 17) association between period of adoption and gender emerged as significant (i = 176 df=3) Female-headed households tended to lag behind male-headed households in adopting row planting Male furmers started adopting the technology as early as 1970 and about half of those who adopted did so before 1980 The majority of female-headed households adopted row planting very recently (1986-90) and there has however been a gradual and consistent increase since 1975
Regarding spacing between and within hills none of the sampled furmers followed exactly the recommended levels but the majority of the furmers (53 ) who used 30 x 90 cm (Table 16) were very close to the recommended spacing The seeding rate commonly used was 1-2 seeds per hill The implication of not using the recommended spacing was that the optimum plaftt density of 44444 plants per hectare was not achieved
Intercropping and crop rotation
Maize was largely grown as a sole crop Thirty-five and five percent of male and female-headed households respectively intercropped maize (Table 18) As a result the i test showed a highly significant association (i=281 df=2) between gender and intercropping
24
Table 18 Intercropping maize with other crops by gender of household head
IntercrOQQed maize Gender Yes No Total
Male 22 (35) 41 (65) 63 Female 2 (5) 35 (95) 37 Total 24 (24) 76 (76) 100
Beans and Potato were used in intercropping with maize but the most common crop used with maize is beans (79 of those who intercropped) No female-headed household intercropped maize with potato The most important reasons for intercropping were land scarcity and increasing food security base
Unlike intercropping most female-headed households practiced crop rotation About 20 of male and 78 of female-headed households rotated maize with beans The rest did not practice crop rotation as they planted maize in the same plots as during the 199192 cropping season
Improved ~rieties
Seventy-six percent of funners responded as having grown improved varieties at one time or another (Table 19) However there was a strong association between gender and the habit of growing improved variety (t= 1228 df= 1) There was a strong indication that male-headed fumilies had adopted the improved variety at greater rate than the female ones Improved varieties ever grown in the swvey area included H614 SR52MalawiZimbabwe KenyaNjombe H6302 Kito and Katumani
Table 19 Growing of improved maize variety by gender of household head
Ever grown Gender No Yes Total
Male 10 (16) 53 (84) 63 (100) Female 14 (38) 23 (62) 37 (100) Total 24 (24) 76 (76) 100 (100)
Although there seems to be association between gender and varieties grown the result was not significant at 5 probability level There was however a tendency that a larger number of male furmers grew variety H614 than female funners and
25
a larger number of female heads have grown variety SR52MalawiZimbabwe than male furmers The most popular variety in the region was H614 as it vvas grown by 88 of those fiumers who had ever grown improved varieties (Table 20) Varieties such as SR52MalawiZimbabwe KenyaNjombe Kito H6302 and Katumani are hardly grown Only one female-headed household has ever grown H6302 and Kito and only one male-headed household had grown Katumani
Table 20 Improved maize varieties ever grown by gender of household head
Variety
Gender H614 SR52 KenyaNjombe Others Total
Male 49 (92) 1 (2) 2 (4) 1 (2) 53 Female 18 (78) 2 (9) 1 (4) 2 (9) 23 Total 67 (88) 3 (4) 3 (4) 3 (4) 76
During the survey year 26 and 74 of respondents grew H6l4 and the local variety respectively However recent surveys have shown that it is difficult to find pure local varieties in the Southern Highlands (Nalitolela 1990) Hence what may be considered as local varieties now were once improved varieties which have undergone genetic deterioration as a result of seed recycling for many years
Table 21 shows periods when furmers started growing improved maize varieties by gender of household head Major adoption of the improved varieties started around 1976 In fuct there is an indication that these varieties were even known by some of furmers back in 1970 as six male-and tMl female-headed households started growing such varieties during 1970-1976 and only three male and one female fumilies started before 1970 Most fiumers adopted improved varieties between 1981 and 1990 while the period from 1986 to 1990 saw a high adoption rate of these varieties Among furmers reported as having grown improved varieties six male- and tMl female-headed households did not indicate the period when they started growing such varieties
26
Table 21 Period when farmers started growing improved maize variety by gender of household head
Period Male Female Total
1976-80 7 (15) 3 (14) 10 (15) 1981-85 10 (21) 4 (19) 14 (20) 1986-90 23 (49) 13 (62) 36 (53) 1991-93 7 (15) 1 (5) 8 (12) Total 47 21 68
Disadoption of Improved Varieties
The majority of those farmers who had adopted improved varieties (64 and 74 for male and female-headed households respectively) reported as having discarded the improved varieties at least once in their lire time The rate of disadoption was however the same among male and remale-headed households (i test is not significant) Most of those farmers (88 82 for male and female-headed households respectively) who discarded improved varieties discarded them between 1986 and 1993 (Table 22) Since reasons fur disadoption was almost similar among both male and female-headed households there was no association between gender and year of disadoption (i=36 df=3) It seems like similar proportion of both male and female farmers discarded improved varieties every year
Table 22 Period improved maize variety was discarded by gender of household head
Period Male Female Total
1976-80 1 (3) 1 (6) 2 1981-85 3 (9) 2 (12) 5 1986-90 14 (41) 8 (47) 22 1991-93 16 (47) 6 (35) 22 Total 34 (67) 17 (33) 51
All improved varieties known to fanners or ever grown in the study area have been discarded at one time or another Among the fanners who had adopted H614 variety about 55 and 66 male- and female-headed households respectively reported as having discarded it
The main reasons for discarding improved varieties were that the varieties were not available (35 of male and 59 female-headed households) expensive to purchase (41 of male and 35 female-headed households) poor taste (7 of male- and
27
no female-headed households) poor milling quality (6 of male- and no female-headed households) cob rot and poor growth (only one male respondent each)
H614 has been discarded mainly for its unavailability (31 of those who have discarded it) and cost (37 of those who have discarded it) Clearly unavailability and cost were the major reasons for the high rate of disadoption
Twenty (out of 21) male and four (out of 5) female-headed households who preferred variety H614 preferred it for its high yielding capability and one respondent from each gender preferred the variety for its taste It seemed that preference was associated with gender (il= 1004 df= 1) male-headed households preferring variety H614 more than female-headed fumilies Most of those farmers specially male ones who responded as having discarded this variety earlier for various reasons still preferred it fur future planting
Weeding
All furmers weeded maize plots at least once during the crop season About 27 of female-headed households weeded their maize plot between September and October during the period of land preparation for some farmers (Thble 23) and a considerable proportion (35) weeded during planting period (Novemshyber-December) About 22 of male-headed households weeded maize during planting (September-October) No male-headed household weeded maize during the major planting period (July-August) Similar proportion of male (37 ) and female-headed households (35 ) weeded maize plot in January a few weeks after maize was planted Weeding dates were however significantly associated with gender (~=116 df=6) greater proportion of female farmers (62) than male farmers (50) weeded their maize land before and including December Greater proportion of female-headed households than male-headed households weeded maize in October while greater proportion of male-headed households than female-headed households weeded their maize field in February
In wet season planting areas the date of first weeding was determined by the date when maize was planted while in the dry season planting areas first weeding was delayed a bit because weed growth is suppressed by low moisture conditions
28
Table 23 Date of first weeding of plot one by gender of household head
Months Male Female Total
September 8 (13) 3 (8) 11 October 6 (9) 7 (19) 13 November 10(16) 7 (19) 17 December 8 (13) 6 (16) 14 January 23 (37) 13 (35) 36 February 8 (13) 1 (3) 9 Total 63 37 100
Weeding of plots 2 3 and 4 were performed by most farmers between September and the last week of November There were no indications of diflerences in weeding date of these plots between the male and female farmers
About 65 male and 57 female-headed households weeded their first plot twice More than 50 of those farmers who had 2nd and 3rd plots did not weed them twice and those having 4th plot did not weed it at all Most of those who undertook second weeding perfurmed it between November and February There was a general tendency that the less number of plots the farmer had the more frequent it was weeded
The majority of fanners (87 male and 95 female-headed households) used hand hoe for weeding Only 5 male and none of the female-headed households used herbicide and about 8 and 5 of male and female- headed households used both hand hoe and herbicides About 67 of those who used herbicide transported it from a distance of more than 4 km The reasons fur not using herbicide are shown in Table 24 It would appear that the lack of knowledge on how to use herbicide and shortage of money to purchase it were the main reasons why farmers did not use herbicide In general a significant association between reasons fur not using herbicide and gender was detected (i=511 df=3) More proportion of female-headed households than the male ones were unable to purchase herbicide for the lack of money and more proportion of male headed- households than the female ones responded that they did not know how to use herbicides
29
Table 24 Reasons for not using herbicide by gender of household head
Reasons Male Female Total
Too expensive 5 (9) 2 (6) 7 (8) Not available 4 (7) 3 (8) 7 (8) No knowledge 31 (55) 15 (42) 46 (50) No money 16 (29) 16 (44) 32 (34) Total 56 (61) 36 (39) 92 (100)
Soil Fertility Management
Available statistics in Tanzania show that the Southern Highlands zone is the major fertilizer consumer in the country For instance between 1989 and 1991 the zone consumed 68 while Mbeya region alone consumed 40504 tones (~O) out of 207406 tones during the same period (URT 1992)
It was found that turners used both organic and inorganic fertilizers As a matter of mct by fertilizer furmers understood inorganic fertilizer For this reason exactly the same response was obtained for the tv) questions apply fertilizer this season and use inorganic fertilizer Consequently 78 female and 46 male-headed households responded yes to the tv) questions indicated above Therefore there was a strong indication that use of inorganic fertilizer was highly associated with gender (y = 21 7 df= 1) female-headed households mvoring the use of inorganic fertilizer more than male-headed households
The major reasons for not using fertilizer included lack of money to purchase fertilizers the soil was already fertile and unavailability of fertilizer in the market Among these reasons unavailability seemed the most serious problem in the survey area (Table 25) A study conducted in Njombe district in lringa region (also in the Southern Highlands) found the same reasons as to why furmers were not using fertilizers (Mwakyembe et al 1992)
Table 25 Reasons for not using fertilizers by gender of household head
Reasons Male Female Total
No money 10 (29) 2 (25) 12 Fertile soil 8 (24) - (0) 8 Not available 16 (47) 6 (75) 22
30
The Chi-square test shows that there is a significant association between gender and reasons for not using fertilizers (~=3072 df=2) Many the female-headed households who were not using fertilizer did so for the reason that fertilizer was not available while more male-headed households than female-headed households did not use fertilizer claiming that their land was already fertile
More male-headed households used organic fertilizer than female farmers Hence there is a significant association between gender and the use of organic fertilizer (~= 18754 df= 1) Fifty-fi~ percent and 25 of male and female farmers used organic fertilizer A large proportion of households used organic fertilizer rather than FYM Mean distance for transporting organic fertilizer from home to the field was 038 Ian (SE=014) the maximum and the minimum distances being ten and zero kilometers respectively The majority (82 ) of households however did not use or transport (7 did not transport) organic fertilizer 10 transported one Ian and only 2 transported more than seven Ian However it seems that male-headed households did not transport fertilizer or transported very short distances as mean distance transported was only 008 Ian (SE=004) while female farmers transported longer distances than male farmers (mean=087 SE=035) When comparing male and female furmers however the t-test lies on the border (t=224) Hence for the reason that there was large variation in distances travelled by female farmers it can be said that there was a moderate difference between male and female farmers Probably this was why most female-headed households tend to use inorganic fertilizer
The majority of farmers (86 ) did not use FYM about 10 used under three bags of manure 3 used between three and seven bags and only one furmer used more than ten bags The mean bags used were 077 (SE=02) The mean bags of FYM used by male- and female-headed households was 082 (SE=025) and 07 (SE =023) respectively However no significant difference was detected between male and female farmers Green manure and plant residues were not used by all the sampled furmers Other studies in the Southern Highlands have shown that to minimize costs furmers do not apply organic fertilizers alongside inorganic fertilizers (ASSP 1992)
Among those furmers that had ever used inorganic fertilizer the majority (74) adopted the technology between the years 1976 and 1990 (Table 26) More male-headed households adopted the technology earlier than female farmers The adoption rate for male-headed households was similar before and after the period 1981-85 However the rate has been increasing but sharply decreased between 1991 and 1993 among the female-headed households probably following the removal of subsidy on fertilizer during that period
31
Table 26 Period when farmers started to use inorganic fertilizer by gender of household head
Period Male Female Total
Before 1970 5 (17) - (0) 5 (9) 1971 - 75 2 (7) 4 (14) 6 (10) 1976-80 5 (17) 6 (17) 10 (17) 1981-85 9 (31) 9 (31) 18 (31) 1986 - 90 5 (17) 10 (34) 15 (26) 1991 - 93 3 (10) 1 (4) 4 (7)
Female-headed households were the major users of fertilizer both as a basal and first top dressing Seventy-three percent female furmers as compared to 25 male furmers used TSP as a basal fertilizer and the same proportion of female furmers as compared to 41 male furmers used either CAN or urea fur first top dressing
There was a strong association between gender and type of first top dressing fertilizer used (x2=883 df= 1) Most male-headed households tended to use CAN (62 ) while most female-headed fumilies tended to use urea (59 )
Mean basal fertilizer used was 064 bags of 50 kg Mean for male- and femaleshyheaded households was 034 bags (SE=009) and 114 bags (SE=016) respectiveshyly The maximum used was 3 bags and the minimum 05 with 95 of all furmers using at most 2 bags There is a significant difference between male and female furmers (t=43) in the use of TSP as a basal fertilizer
Mean first top dressing fertilizer used by all furmers was 081 bags (SE =01) with a maximum of 4 bags and a minimum of 05 bags the mean for male and femaleshyheaded households being 052 (SE=0105) and 128 (SE=018) respectively Female-headed households used significantly greater amount of fertilizer (t=36) for first top dressing
Almost similar proportion of male (19 ) and female (22 ) headed households used fertilizer fur second top dressing Equal proportion of male purchased CAN and urea whereas the majority of female-headed households who used fertilizer fur second top dressing purchased CAN
The total money spent on basal fertilizer (TSP) by all farmers is Tshs 136570 (US$ 24831) Mean spent by male and female-headed households being Tshs 6984 (US$ 127) (SE=1712) and 2502 (US$ 455) (SE=3436) respectively The majority of furmers (77) spent between Tshs 2200 (US$ 400) and Tshs 4400 (US$ 800) to purchase fertilizer The money spent by all farmers ranged from Tshs 6600 (US$ 1200) to 12100 (US$ 2200) The amount of money spent on TSP by
32
female-headed households was more than that of the male ones (t=47) Similarly about Tshs 193000 (US$ 35100) was spent by all furmers on purchasing fertilizer for first top dressing of which Tshs 91000 (US$ 16545) was spent on purchase of CAN and the remaining Tshs 102000 (US$ 18545) on urea Mean spent by male and female-headed households was Tshs 29375 (US$ 534) (SE=5281) and Tshs 4000 (US$ 737) (SE=674) for CAN while that spent on urea was Tshs 2550 (US$ 464) (SE=3905) and Tshs 47812 (US$ 869) (SE=4978) respectively
Those male and female households who purchased fertilizer for first top-dressing purchased on average 15 and 2 bags of CAN and 085 and 16 bags of urea respectively This means that nearly every female-headed household purchased twice as much urea as that of male-headed household
Other forms of soil fertility management in the survey area included land fullowing (Table 12) and crop rotation especially rotation of maize with beans and potatoes Thirty-three percent of the sampled furmers claimed that they burnt maize stover to fertilize the soil
All farmers travelled at most one Ian to obtain organic fertilizer while the majority of furmers who used inorganic fertilizer travelled at least 3 Ian (Thb1e 27) This was because inorganic fertilizers were usually sold and distributed in the nearby towns hence distances travelled by a farmer may therefore reflect hisher proximity to the urban center
Like female-headed households who usually travelled longer distances than male-headed ones to obtain FYM male farmers travelled longer distances to get inorganic fertilizer Consequently about 59 and 70 of female- and male-headed households who used inorganic fertilizer travelled at least 3 Ian to obtain it
Apart from the lack of money and fertilizer availability the fact that some of the farmers had to travel long distances to obtain inorganic fertilizers might be an additional bottleneck in the use of inorganic fertilizers Furthermore the fact that the government has removed subsidy on fertilizers more and more farmers will abandon the use of fertilizer and this will have adverse implications on maize industry and food security in the region and the nation at large Nevertheless the removal of subsidy may also promote organic farming in the study area because fanners already know the importance of soil fertility management
33
Table 27 Distances travelled to obtain organic and inorganic fertilizers by gender of household head
Distances (km)
Fertilizer Gender Below 1 1-2 3-4 Over 4 Total
Male 15 15 Organic Female 9 9
Total 24 24
Male 9 3 11 18 41 Inorganic Female 3 4 6 4 17
Total 12 7 17 22 58
Pests and Disease Control
The most important pests and diseases were outlined under the section that described the study area MSV was the main disease in the survey area Farmers are advised to grow an improved variety TMV-l which has proved to resist MSV attack Those who cannot afford to buy new seeds of TMV-l are advised to plant early in order to avoid massive build up of the vectors
Farmers in the survey area use maize stover as animal feed or burn it to kill pests Eighty-one percent of male-headed households and ninety-seven percent of the female ones burnt their maize stover to clear the field or fertilize the soil or eliminate pests Female-headed households tended to burn maize stover rather than use it as animal feed (~=134 df=l) because few of them keep cattle On one hand larger proportion of female-headed households (78 ) than the male ones (70 ) reported their maize crop having been attacked by stalk borer on the other hand greater proportion of the crop areas of male-headed households had been affected by pests than for the female ones That is even if more proportion of feshymale-headed households complained about the damage caused by pests their total crop area affected was less than that of the male ones probably because on the average they have small maize crop area as compared to male-headed households
About 53 of male-headed households and 47 of female-headed households reported that only one plot of their furm had been affected by disease Twenty-two percent of male and 8 of female headed households had two plots of their maize area affected by disease The fuct that 73 of the furmers reported that at least one of their maize plots was affected by disease indicated the extent of the disease problem in the survey area Regarding protective measures out of those furmers whose area was attacked 77 of male furmers and 86 of female furmers responded that they did not do anything to control the situation when their crop area
34
was attacked by pestsdiseases Among furmers experiencing pestdisease problems in their maize plots 81 male- and 83 female-headed households indicated that their local maize variety was the most affected one
Harvesting and Post-harvest Storage System
The majority of male-headed households (75 ) harvested their first maize plots between April and June whereas 65 of female-headed households harvested in May and June Since most female-headed households planted their first maize plot later than the male ones they also tended to harvest one month later Most male-headed households who have second plots of maize harvested them between April and May while the female ones harvested between May and July The remaining two plots also tended to be harvested by both male and female furmers between April and June It can be concluded therefore that the main harvesting period for all furmers was between April and June
Nearly half of the respondents stored their maize shelled in bags and about 72 of them protected their maize grain from damage using Actellic dust The next common storage system was storing unshelled maize in Vihenge (a local storage structure
which is made of mud bricks or twigs and then plastered with mud or roN dung) followed by spreading on roof (Table 28) The least used method is storing shelled maize in Vihenge (used by about 11 of furmers) Storage methods showed a significant association with gender (~= 1225 df= 1) That is male-headed households seemed to fuvor storing shelled maize in Vihenge while female-headed households preferred spreading maize on roofS
Table 28 Maize storage systems by gender of household head
Storage Male Female Total
Shelled in bags 31 (49) 18 (49) 49 Shelled in vihenge 10 (16) 1 (3) 11 Unshelled in vihenge 14 (22) 9 (24) 23 Spread on roof 8 (13) 9 (24) 17 Total 63 37 100
In respect to protection methods funners either stored their maize without treating or treating it with pyrethrum or Actellic dust Actellic dust treatment was used by the majority of funners 71 of male-headed and 73 of the female-headed households Only 8 male-headed households and 5 the female ones treated their maize with pyrethrum flowers There was no association between gender and protection method (x2 =O74 df=I)
35
Factors Affecting Adoption
Farmers in Mbeya region are at various levels of adopting various components of the recommended package of improved maize production Besides furmers have adopted these components gradually
The four major factors that contributed to this gradual adoption are cost of technologies environmental factors timely availability of inputs and source of infurmation on new technologies
Cost of the lechnology
Technologies which require little cash outlay such as row planting are easily taken up by furmers Row planting was adopted by almost all sampled furmers (96 ) mainly because it was less costly and had an added advantage of simplifying weeding Farmers who did not grow improved seed during the 199293 cropping season mentioned high cost (22 ) and shortage of cash to buy it (54 ) as their reasons for not growing improved seed
Thirty-eight percent of the sampled furmers have over the years discarded or disadopted improved seeds because they were too expensive Also 38 and 21 of the furmers did not use herbicides and fertilizer respectively as they were expensive
Environmental Factors
Environmental stress constrained the adoption of some of the recommendations especially where maize is planted during the dry season which utilizes residual moisture in the soil Farmers who dry planted their maize did not apply basal fertilizer This might be because of the fear of scotching their maize seed due to low soil moisture Others did not perfurm the second weeding apparently because vigorous weed germination will be suppressed by the moisture conditions Prevalence of pests and diseases especially stalk borer and MSV might have affected the adoption of improved maize varieties
Timely Availability of Inputs
Lack andor timely availability of inputs were widely cited as constraints to using them Forty-seven percent of male-headed and 75 of female-headed households cited unavailability of fertilizer as the main reason why they were not using it The
36
unavailability of improved maize seed was considered a bottleneck to its use For instance 55 of male-headed and 66 of female-headed households who had discarded the use of variety H614 cited unavailability of seed as the main reason Other studies in Southern Highlands also suggested unavailability of inputs as major constraint to their use (Lyimo and Temu 1992) Another problem faced by some turners in the survey area in respect to availability of inputs was the assumption by input suppliers that maize in the Southern Highlands is planted only during the wet season and this is the time when they tend to make the inputs available
Sources of Information on New Technologies
About 46 male-headed and 57 of female-headed households had never received a visit from extension service Both male and female funners had received visits in almost equal proportion and there was no significant difrerence between them
Extension workers and other furmers were found to be the major sources of infurmation on hybrid maize seeds row planting fertilizer use and the use of herbicides Table 29 shows the source of information fur obtaining hybrid seeds by gender of household NGOs and traders were not important sources of information on new technologies as seen in the case of hybrid maize seeds In general source of infurmation were not associated with gender (~=538 df=4) This means that all sources of infurmation that were at funners disposal did not difrerentiate between gender
Given that the extension service is charged with the responsibility of extending infurmation on new technologies their low rates of contact with funners may be acting as a constraint to the use of these technologies For instance only 27 of male-headed and 32 of female-headed households had been visited by extension workers one month before the survey of the sampled furmers fur this study Experience with Sasakawa-Global 2000 shows that with close extension interaction furmers can increase their crop yields tremendously (Quinones et al 1992)
Table 29 Sources of information for obtaining hybrid maize seeds by gender of household head
Source Male Female Total
Extension agents 22 (35) 14 (38) 36 NGOs 7 (11) 2 (5) 9 Traders 1 (2) 1 (3) 2 Other farmers 23 (37) 11 (30) 34 No source 10 (15) 9 (24) 19
37
Conclusions and Implications for Research Extension
and Policy
Farmers in the survey area are generally aware of the recommended maize production technologies However the use of these technologies are limited by factors such as cost environmental stresses lack and or timely availability and lack of infurmation
The problem of stalk borer needs to be addressed by research perhaps by developing tolerant maize varieties to this pest Research has addressed the problem of MSV by developing the variety TMV-l that is supposed to be resistant to this virus although farmers are not generally aware of this variety The range of hybrids maize that are acceptable and affordable by farmers need to be widened At present it would appear that H614 is the only available and acceptable hybrid maize And even for H6l4 the dis adoption rate was very high mainly due to its unavailability and high cost which is a reflection of the status of the maize seed industry in the country
There is agreat need for extension service to improve its contact with farmers The current low contact rate is a major constraint to the use of some of the recommendshyed technologies Lack of lmowledge on how to use herbicides for example limits the adoption of this technology especially by male-headed households while the lack of infurmation on such technology as variety TMV-l has limited its use This might also be a reflection of the lack of research-extension linkage
Other studies in Tanzania and the Southern Highlands in particular have called for strengthening this linkage (Keregero 1992 Ekpere and Shetto 1992) Given the constraints faced by the extension service in meeting the needs of individual farmers extending messages to a group of farmers rather than individuals should be strengthened
The cost of technology is a major constraint to its adoption The elimination of subsidy on price of seed and fertilizer due to structurdl adjustment programs will certainly tighten this constraint The subsidy issue needs to be addressed very carefully by policy-makers because given the low adoption rates of seed-fertilizer technology and its critical role to increasing maize production which is very important for Tanzanias food security there might be a case for maintaining seedshyfertilizer subsidy in the short-to medium-terms
38
Policy-makers should address the possibility of providing appropriate credit for seed and fertilizer to small-scale maize producers The importance of credit to increasing maize production has been amply demonstrated by SO-2000 (Quinnones et al 1992) Credit will also need to be targeted to female-headed households For instance 44 of female-headed households could not use herbicides due to the lack of money to purchase it compared to 29 of male-headed households
Lack andor timely availability of technology were the major constraint to adoption of technology thus policy-makers should create a favorable environment for the provision of these inputs especially seed and fertilizer by facilitating the developshyment of a viable private sector In the short-to medium-terms the provision of credit to traders to purchase seed and fertilizer and also to construct storage facilities will assist in promoting the development of the private sector In the long-term infrastructural development such as access roads and local markets will be needed especially to reduce the long distances travelled by furmers to obtain these inputs Other alternatives such as giving researchers at Uyole the mandate to produce seed and sell to farmers should be explored
Policy-makers should examine the imbalance between male and female furmers in terms of access to production resources such as land and capital as well as to infurmation on new technologies For instance male-headed households had more access to land than female-headed households Ninety-six percent of male-headed households had some formal education while 82 of female-headed households had no formal education
It was also observed that more female-headed households faced the problem of availability of seed (59 ) and fertilizer (75 ) compared to male-headed households ( 35 and 47 respectively) This certainly calls for a more concerted effort by policy-makers to target female-headed households not only on equity grounds but also to assist in increased maize production which will enhance the nations food security
39
References
Agricultural Sector Support Programme (ASSP) 1992 Baseline survey of the Southern Highlands of Thnzania Food Security Unit Ministry of Agriculture Dar es Salaam Th nzanil
Croon I J Deutsch AEM Temu 1984 Maize production in Thnzania s Southern Highlands Current Status and Recommendations for the Future CIMMYT Mexico
Cooper JPM and Law R 1976 The effect and Importance of soil temperature in determining the early growth vigour and final yields of hybrid maize in the Highlands of Kenya A paper presented at the World Meteorological Organisation symposium on the agriclimatology of maize July 5-9 1976 Ames Iowa
Ekpere JA and MC Shetto 1992 Research-extension Linkages and Service to the Small Farmer in the Southern Highlands of Thnzania Evidence from Uyole Agricultural Center Mbeya In Ekpere JA D J Rees RP Mbwile and N G Lyimo (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and future prospects 5-9 October 1992 Mbeya Thnzania
Keregero KJB 1992 Transferring Agricultural Technologies to Farmenl lbwards Stronger Research-Extension Linkages In Ekpere JA Rees DJ Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achieveshyments and future prospects 5-9 October 1992 Mbeya Thnzania
Lyimo NG and Temu AEM 1992 The Southern Highlands Maize Improvement Programme Achievements and strategies for future research In Ekpere JA Rees DJ Mbwile RP and NJ Lyimo (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Marandu wYP Lyimo NG Thmu AEM and D Kabungo 1989 Maize Improveshyment in the Southern Highlands of Thnzania In Gebrekidan B(ed) Maize Improvement Production and Protection in Eastern and Southern African Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Matowo PR and Mgema wG 1988 Optimum weeding in maize in the low and Intermediate areas of Thnzania In Moshi AT and Ransom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mayagilla AH 1988 Opening address In Moshi A I and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mduruma ZO Moshi AJ and PR Matowo 1988 Use of Carbofuran as a seed dressing insecticide for reducing incidence of streak virus disease in maize In Moshi AJ and Ransom JK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Mguni C 1988 Effect of application methods of insecticide on maize streak virus disease in maize In Moshi AJ and Rasnom JK (eds) Maize Research in Thnzania Proceedings of the Finlt Thnzania National Maize Research Workshop held at Arusha TARO Dar es Salaam Thnzania
Moshi AJ and Nnko SN 1989 Maize Seed Production and utilization in Thnzania In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern
40
Africa Proceedings of the 3rd Eastern and Southern Africa Regional Maize workshop 18shy22 September 1989 Nairobi Kenya
Mussei ANK and IK Shiyumbi 1992 Review of studies of the farming systems of the Southern Highlands of Thnzania 1970-1990 In Ekpere IA Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Reshysearch Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Mwakyembe CMA Rees 01 and AE Simfulcwe 1992 A description of farm enterprise combinations and production practices in the Southern Highlands of Thnzania In Ekpere la Rees 01 Mbwile RP and Lyimo NG (eds) Proceedings of an International Conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Future Prospects 5-9 October 1992 Mbeya Thnzania
Nalitolela Al 1990 A Diagnostic survey report of the Uporoto - Umalila highlands of Mbeya region Uyole Agricultural Centre Mbeya Thnzania
Ngwira P 1989 The status of maize disease in Malawi In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Ochor TE Kedera Cl and Ochieng 1989 Effects of Delayed Harvest and Host Genotype on the Incidence of Ear Rots in Western Kenya In Gebrekidan B (ed) Maize Improvement Production and Protection in Eastern and Southern Africa Proceedings of the 3rd Eastern and Southern African Regional Maize workshop 18-22 September 1989 Nairobi Kenya
Prasad R 1978 Management practices for improving maize yields Technology for Increasing Food Production FAO Rome
Quinones M Foster AM Akibo-Betts D and NP Siciliana 1992 Transferring Improved Production Technologies to small-scale farmers Methodology used by SasakawashyGlobal 2000 in Africa In Ekpere la Rees Dl Mbwile RP and Lyimo NG (eds) Proceedings of an International conference on Agricultural Research Training and Technology Transfer in the Southern Highlands of Thnzania Past Achievements and Suture Prospects 5-9 October 1992 Mbeya Thnzania
Semuguruka GH 1988 Maize Research in Thnzania Past Present and Future In Moshi A1 and Ransom lK (eds) Maize Research in Thnzania Proceedings of the First Thnzania National Maize Research Workshop Arusha TARO Dar es Salaam Thnzania
Temu AEM 1991 Kilimo Bora cha Mahindi Extension leaflet No 35 Uyole Agricultural Centre
United Republic of Thnzania (URT) 1992 Agmultural Statistics for 1989 Dar es Salaam Thnzania
Uyole Agricultural Centre (UAC) 1974175 - 1991192 Annual reports Mbeya Thnzania
41
ISBN 92- 9146--013--3 Pmted lit LA Addis Ababa Ethiopia
